| Title | Date | Abstract | Comment |
|---|---|---|---|
| FA-Seg: A Fast and Accurate Diffusion-Based Method for Open-Vocabulary Segmentation | 2026-04-29 | ShowOpen-vocabulary semantic segmentation (OVSS) aims to segment objects from arbitrary text categories without requiring densely annotated datasets. Although contrastive learning based models enable zero-shot segmentation, they often lose fine spatial precision at pixel level, due to global representation bias. In contrast, diffusion-based models naturally encode fine-grained spatial features via attention mechanisms that capture both global context and local details. However, they often face challenges in balancing the computation costs and the quality of the segmentation mask. In this work, we present FA-Seg, a Fast and Accurate training-free framework for open-vocabulary segmentation based on diffusion models. FA-Seg performs segmentation using only a (1+1)-step from a pretrained diffusion model. Moreover, instead of running multiple times for different classes, FA-Seg performs segmentation for all classes at once. To further enhance the segmentation quality, FA-Seg introduces three key components: (i) a dual-prompt mechanism for discriminative, class-aware attention extraction, (ii) a Hierarchical Attention Refinement Method (HARD) that enhances semantic precision via multi-resolution attention fusion, and (iii) a Test-Time Flipping (TTF) scheme designed to improve spatial consistency. Extensive experiments show that FA-Seg achieves state-of-the-art training-free performance, obtaining 43.8% average mIoU across PASCAL VOC, PASCAL Context, and COCO Object benchmarks while maintaining superior inference efficiency. Our results demonstrate that FA-Seg provides a strong foundation for extendability, bridging the gap between segmentation quality and inference efficiency. The source code is available at https://github.com/chequanghuy/FA-Seg. |
|
| Seeking Consensus: Geometric-Semantic On-the-Fly Recalibration for Open-Vocabulary Remote Sensing Semantic Segmentation | 2026-04-29 | ShowOpen-vocabulary semantic segmentation (OVSS) in remote sensing images is a promising task that employs textual descriptions for identifying undefined land cover categories. Despite notable advances, existing methods typically employ a static inference paradigm, overlooking the distinct distribution of each scene, resulting in semantic ambiguity in diverse land covers and incomplete foreground activation. Motivated by this, we propose Seeking Consensus, termed SeeCo, a plug-and-play framework to boost the performance of training-free OVSS models in remote sensing images, which recalibrates arbitrary OVSS models on-the-fly by seeking dual consensus: geometric consensus learning (GCL) through multi-view consistent observations and semantic consensus learning (SCL) via textual description adaptive calibration, which assists collaborative recalibration of visual and textual semantics. The two consensus are injected via an online consensus injector (OCI), effectively alleviating the under-activation and semantic bias. SeeCo requires no specific training process, yet recalibrates semantic-geometric alignment for each unique scene during inference. Extensive experiments on eight remote sensing OVSS benchmarks show consistent gains, proving its effectiveness and universality. |
11 pages, 9 figures |
| DouC: Dual-Branch CLIP for Training-Free Open-Vocabulary Segmentation | 2026-04-27 | ShowOpen-vocabulary semantic segmentation requires assigning pixel-level semantic labels while supporting an open and unrestricted set of categories. Training-free CLIP-based approaches preserve strong zero-shot generalization but typically rely on a single inference mechanism, limiting their ability to jointly address unreliable local tokens and insufficient spatial coherence. We propose DouC, a training-free dual-branch CLIP framework that decomposes dense prediction into two complementary components. OG-CLIP improves patch-level reliability via lightweight, inference-time token gating, while FADE-CLIP injects external structural priors through proxy attention guided by frozen vision foundation models. The two branches are fused at the logit level, enabling local token reliability and structure-aware patch interactions to jointly influence final predictions, with optional instance-aware correction applied as post-processing. DouC introduces no additional learnable parameters, requires no retraining, and preserves CLIP's zero-shot generalization. Extensive experiments across eight benchmarks and multiple CLIP backbones demonstrate that DouC consistently outperforms prior training-free methods and scales favorably with model capacity. |
|
| Open-Vocabulary Semantic Segmentation Network Integrating Object-Level Label and Scene-Level Semantic Features for Multimodal Remote Sensing Images | 2026-04-27 | ShowSemantic segmentation of multi-modal remote sensing imagery plays a pivotal role in land use/land cover (LULC) mapping, environmental monitoring, and precision earth observation. Current multi-modal approaches mainly focus on integrating complementary visual modalities, yet neglect the incorporating of non-visual textual data - a rich source of knowledge that can bridge semantic gaps between visual patterns and real-world concepts. To address this limitation, we propose TSMNet, a text supervised multi-modal open vocabulary semantic segmentation network that synergistically integrates textual supervision with visual representation for open-vocabulary semantic segmentation. Unlike conventional multi-modal segmentation frameworks, TSMNet introduces a dual-branch text encoder to extract both scene-level semantic and object-level label information from various textual data, enabling dynamic cross-modal fusion. These text-derived features dynamically interact with visual embeddings through the proposed text-guided visual semantic fusion module, enabling domain-aware feature refinement and human-interpretable decision-making. To verify our method, we innovatively construct two new multi-modal datasets, and carry out extensive experiments to make a comprehensive comparison between the proposed method and other state-of-the-art (SOTA) semantic segmentation models. Results demonstrate that TSMNet achieves superior segmentation accuracy while exhibiting robust generalization capabilities across diverse geographical and sensor-specific scenarios. This work establishes a new paradigm for explainable remote sensing analysis, demonstrating that textual knowledge integration significantly enhances model generalizability. The source code will be available at https://github.com/yeyuanxin110/TSMNet |
|
| SegEarth-OV3: Exploring SAM 3 for Open-Vocabulary Semantic Segmentation in Remote Sensing Images | 2026-04-22 | ShowMost existing methods for training-free open-vocabulary semantic segmentation are based on CLIP. While these approaches have made progress, they often face challenges in precise localization or require complex pipelines to combine separate modules, especially in remote sensing scenarios where numerous dense and small targets are present. Recently, Segment Anything Model 3 (SAM 3) was proposed, unifying segmentation and recognition in a promptable framework. In this paper, we present a comprehensive exploration of applying SAM 3 to the remote sensing open-vocabulary tasks (i.e., 2D semantic segmentation, change detection, and 3D semantic segmentation) without any training. First, we implement a mask fusion strategy that combines the outputs from SAM 3's semantic segmentation head and the Transformer decoder (instance head). This allows us to leverage the strengths of both heads for better land coverage. Second, we utilize the presence score from the presence head to filter out categories that do not exist in the scene, reducing false positives caused by the vast vocabulary sizes and patch-level processing in geospatial scenes. Furthermore, we extend our method to open-vocabulary change detection by a joint instance- and pixel-level verification strategy built directly upon our fused logits. We evaluate our method on extensive remote sensing datasets and tasks, including 20 segmentation datasets, 3 change detection datasets, and a 3D segmentation dataset. Experiments show that our method achieves promising performance, demonstrating the potential of SAM 3 for remote sensing open-vocabulary tasks. Our code is released at https://github.com/earth-insights/SegEarth-OV-3. |
|
| CoCo-SAM3: Harnessing Concept Conflict in Open-Vocabulary Semantic Segmentation | 2026-04-21 | ShowSAM3 advances open-vocabulary semantic segmentation by introducing a prompt-driven mask generation paradigm. However, in multi-class open-vocabulary scenarios, masks generated independently from different category prompts lack a unified and inter-class comparable evidence scale, often resulting in overlapping coverage and unstable competition. Moreover, synonymous expressions of the same concept tend to activate inconsistent semantic and spatial evidence, leading to intra-class drift that exacerbates inter-class conflicts and compromises overall inference stability. To address these issues, we propose CoCo-SAM3 (Concept-Conflict SAM3), which explicitly decouples inference into intra-class enhancement and inter-class competition. Our method first aligns and aggregates evidence from synonymous prompts to strengthen concept consistency. It then performs inter-class competition on a unified comparable scale, enabling direct pixel-wise comparisons among all candidate classes. This mechanism stabilizes multi-class inference and effectively mitigates inter-class conflicts. Without requiring any additional training, CoCo-SAM3 achieves consistent improvements across eight open-vocabulary semantic segmentation benchmarks. |
|
| T-REN: Learning Text-Aligned Region Tokens Improves Dense Vision-Language Alignment and Scalability | 2026-04-20 | ShowDespite recent progress, vision-language encoders struggle with two core limitations: (1) weak alignment between language and dense vision features, which hurts tasks like open-vocabulary semantic segmentation; and (2) high token counts for fine-grained visual representations, which limits scalability to long videos. This work addresses both limitations. We propose T-REN (Text-aligned Region Encoder Network), an efficient encoder that maps visual data to a compact set of text-aligned region-level representations (or region tokens). T-REN achieves this through a lightweight network added on top of a frozen vision backbone, trained to pool patch-level representations within each semantic region into region tokens and align them with region-level text annotations. With only 3.7% additional parameters compared to the vision-language backbone, this design yields substantially stronger dense cross-modal understanding while reducing the token count by orders of magnitude. Specifically, T-REN delivers +5.9 mIoU on ADE20K open-vocabulary segmentation, +18.4% recall on COCO object-level text-image retrieval, +15.6% recall on Ego4D video object localization, and +17.6% mIoU on VSPW video scene parsing, all while reducing token counts by more than 24x for images and 187x for videos compared to the patch-based vision-language backbone. The code and model are available at https://github.com/savya08/T-REN. |
|
| Spatial-Regularization-Aware Dual-Branch Collaborative Inference for Training-Free OVSS in Remote Sensing Imagery | 2026-04-18 | ShowHigh-resolution remote sensing images contain densely distributed objects with pronounced scale variations and complex boundaries, which impose higher demands on both the geometric localization and semantic prediction capabilities of semantic segmentation models. Existing training-free open-vocabulary semantic segmentation (OVSS) methods typically fuse Contrastive Language-Image Pretraining (CLIP) and vision foundation models (VFMs) using one-way injection and shallow post-processing strategies, making it difficult to satisfy these requirements. To address this issue, we propose a spatial-regularization-aware dual-branch collaborative inference framework for training-free OVSS, termed SDCI. First, during feature encoding, SDCI introduces a cross-model attention fusion (CAF) module, which guides collaborative inference by injecting self-attention maps into each other. Second, we propose a bidirectional cross-graph diffusion refinement (BCDR) module that enhances the reliability of dual-branch segmentation scores through iterative random-walk diffusion. Finally, we incorporate low-level superpixel structures and develop a convex-optimization-based superpixel collaborative prediction (CSCP) mechanism to further refine object boundaries. Experiments on multiple remote sensing semantic segmentation benchmarks demonstrate that our method achieves better performance than existing approaches. Our code is available at https://github.com/yu-ni1989/SDCI. |
|
| SENSE: Stereo OpEN Vocabulary SEmantic Segmentation | 2026-04-17 | ShowOpen-vocabulary semantic segmentation enables models to segment objects or image regions beyond fixed class sets, offering flexibility in dynamic environments. However, existing methods often rely on single-view images and struggle with spatial precision, especially under occlusions and near object boundaries. We propose SENSE, the first work on Stereo OpEN Vocabulary SEmantic Segmentation, which leverages stereo vision and vision-language models to enhance open-vocabulary semantic segmentation. By incorporating stereo image pairs, we introduce geometric cues that improve spatial reasoning and segmentation accuracy. Trained on the PhraseStereo dataset, our approach achieves strong performance in phrase-grounded tasks and demonstrates generalization in zero-shot settings. On PhraseStereo, we show a +2.9% improvement in Average Precision over the baseline method and +0.76% over the best competing method. SENSE also provides a relative improvement of +3.5% mIoU on Cityscapes and +18% on KITTI compared to the baseline work. By jointly reasoning over semantics and geometry, SENSE supports accurate scene understanding from natural language, essential for autonomous robots and Intelligent Transportation Systems. |
|
| Seg2Change: Adapting Open-Vocabulary Semantic Segmentation Model for Remote Sensing Change Detection | 2026-04-13 | ShowChange detection is a fundamental task in remote sensing, aiming to quantify the impacts of human activities and ecological dynamics on land-cover changes. Existing change detection methods are limited to predefined classes in training datasets, which constrains their scalability in real-world scenarios. In recent years, numerous advanced open-vocabulary semantic segmentation models have emerged for remote sensing imagery. However, there is still a lack of an effective framework for directly applying these models to open-vocabulary change detection (OVCD), a novel task that integrates vision and language to detect changes across arbitrary categories. To address these challenges, we first construct a category-agnostic change detection dataset, termed CA-CDD. Further, we design a category-agnostic change head to detect the transitions of arbitrary categories and index them to specific classes. Based on them, we propose Seg2Change, an adapter designed to adapt open-vocabulary semantic segmentation models to change detection task. Without bells and whistles, this simple yet effective framework achieves state-of-the-art OVCD performance (+9.52 IoU on WHU-CD and +5.50 mIoU on SECOND). Our code is released at https://github.com/yogurts-sy/Seg2Change. |
21 pages, 15 figures |
| FF3R: Feedforward Feature 3D Reconstruction from Unconstrained views | 2026-04-10 | ShowRecent advances in vision foundation models have revolutionized geometry reconstruction and semantic understanding. Yet, most of the existing approaches treat these capabilities in isolation, leading to redundant pipelines and compounded errors. This paper introduces FF3R, a fully annotation-free feed-forward framework that unifies geometric and semantic reasoning from unconstrained multi-view image sequences. Unlike previous methods, FF3R does not require camera poses, depth maps, or semantic labels, relying solely on rendering supervision for RGB and feature maps, establishing a scalable paradigm for unified 3D reasoning. In addition, we address two critical challenges in feedforward feature reconstruction pipelines, namely global semantic inconsistency and local structural inconsistency, through two key innovations: (i) a Token-wise Fusion Module that enriches geometry tokens with semantic context via cross-attention, and (ii) a Semantic-Geometry Mutual Boosting mechanism combining geometry-guided feature warping for global consistency with semantic-aware voxelization for local coherence. Extensive experiments on ScanNet and DL3DV-10K demonstrate FF3R's superior performance in novel-view synthesis, open-vocabulary semantic segmentation, and depth estimation, with strong generalization to in-the-wild scenarios, paving the way for embodied intelligence systems that demand both spatial and semantic understanding. |
CVPR ...CVPR 2026 Findings. Project Page: https://chaoyizh.github.io/ff3r_project/ |
| OV-Stitcher: A Global Context-Aware Framework for Training-Free Open-Vocabulary Semantic Segmentation | 2026-04-10 | ShowTraining-free open-vocabulary semantic segmentation(TF-OVSS) has recently attracted attention for its ability to perform dense prediction by leveraging the pretrained knowledge of large vision and vision-language models, without requiring additional training. However, due to the limited input resolution of these pretrained encoders, existing TF-OVSS methods commonly adopt a sliding-window strategy that processes cropped sub-images independently. While effective for managing high-resolution inputs, this approach prevents global attention over the full image, leading to fragmented feature representations and limited contextual reasoning. We propose OV-Stitcher, a training-free framework that addresses this limitation by stitching fragmented sub-image features directly within the final encoder block. By reconstructing attention representations from fragmented sub-image features, OV-Stitcher enables global attention within the final encoder block, producing coherent context aggregation and spatially consistent, semantically aligned segmentation maps. Extensive evaluations across eight benchmarks demonstrate that OV-Stitcher establishes a scalable and effective solution for open-vocabulary segmentation, achieving a notable improvement in mean Intersection over Union(mIoU) from 48.7 to 50.7 compared with prior training-free baselines. |
|
| RADSeg: Unleashing Parameter and Compute Efficient Zero-Shot Open-Vocabulary Segmentation Using Agglomerative Models | 2026-04-10 | ShowOpen-vocabulary semantic segmentation (OVSS) underpins many vision and robotics tasks that require generalizable semantic understanding. Existing approaches either rely on limited segmentation training data, which hinders generalization, or apply zero-shot heuristics to vision-language models (e.g CLIP), while the most competitive approaches combine multiple models to improve performance at the cost of high computational and memory demands. In this work, we leverage an overlooked agglomerative vision foundation model, RADIO, to improve zero-shot OVSS along three key axes simultaneously: mIoU, latency, and parameter efficiency. We present the first comprehensive study of RADIO for zero-shot OVSS and enhance its performance through self-correlating recursive attention, self-correlating global aggregation, and computationally efficient RADIO SAM mask refinement. Our approach, RADSeg, achieves 6-30% mIoU improvement in the base ViT class while being 3.95x faster and using 2.5x fewer parameters. Surprisingly, RADSeg-base (106M) outperforms previous combinations of huge vision models (850-1350M) in mIoU, achieving state-of-the-art accuracy with substantially lower computational and memory cost. |
Accep...Accepted to CVPR'26 Findings Code at https://radseg-ovss.github.io/ |
| Direct Segmentation without Logits Optimization for Training-Free Open-Vocabulary Semantic Segmentation | 2026-04-09 | ShowOpen-vocabulary semantic segmentation (OVSS) aims to segment arbitrary category regions in images using open-vocabulary prompts, necessitating that existing methods possess pixel-level vision-language alignment capability. Typically, this capability involves computing the cosine similarity, \ie, logits, between visual and linguistic features, and minimizing the distribution discrepancy between the logits and the ground truth (GT) to generate optimal logits that are subsequently used to construct segmentation maps, yet it depends on time-consuming iterative training or model-specific attention modulation. In this work, we propose a more direct approach that eschews the logits-optimization process by directly deriving an analytic solution for the segmentation map. We posit a key hypothesis: the distribution discrepancy encodes semantic information; specifically, this discrepancy exhibits consistency across patches belonging to the same category but inconsistency across different categories. Based on this hypothesis, we directly utilize the analytic solution of this distribution discrepancy as the semantic maps. In other words, we reformulate the optimization of the distribution discrepancy as deriving its analytic solution, thereby eliminating time-consuming iterative training, freeing us from model-specific attention modulation, and achieving state-of-the-art performance on eight benchmark datasets. |
Accep...Accepted by CVPR 2026 |
| Decouple and Rectify: Semantics-Preserving Structural Enhancement for Open-Vocabulary Remote Sensing Segmentation | 2026-04-02 | ShowOpen-vocabulary semantic segmentation in the remote sensing (RS) field requires both language-aligned recognition and fine-grained spatial delineation. Although CLIP offers robust semantic generalization, its global-aligned visual representations inherently struggle to capture structural details. Recent methods attempt to compensate for this by introducing RS-pretrained DINO features. However, these methods treat CLIP representations as a monolithic semantic space and cannot localize where structural enhancement is required, failing to effectively delineate boundaries while risking the disruption of CLIP's semantic integrity. To address this limitation, we propose DR-Seg, a novel decouple-and-rectify framework in this paper. Our method is motivated by the key observation that CLIP feature channels exhibit distinct functional heterogeneity rather than forming a uniform semantic space. Building on this insight, DR-Seg decouples CLIP features into semantics-dominated and structure-dominated subspaces, enabling targeted structural enhancement by DINO without distorting language-aligned semantics. Subsequently, a prior-driven graph rectification module injects high-fidelity structural priors under DINO guidance to form a refined branch, while an uncertainty-guided adaptive fusion module dynamically integrates this refined branch with the original CLIP branch for final prediction. Comprehensive experiments across eight benchmarks demonstrate that DR-Seg establishes a new state-of-the-art. |
|
| IGLOSS: Image Generation for Lidar Open-vocabulary Semantic Segmentation | 2026-04-01 | ShowThis paper presents a new method for the zero-shot open-vocabulary semantic segmentation (OVSS) of 3D automotive lidar data. To circumvent the recognized image-text modality gap that is intrinsic to approaches based on Vision Language Models (VLMs) such as CLIP, our method relies instead on image generation from text, to create prototype images. Given a 3D network distilled from a 2D Vision Foundation Model (VFM), we then label a point cloud by matching 3D point features with 2D image features of these prototypes. Our method is state-of-the-art for OVSS on nuScenes and SemanticKITTI. Code, pre-trained models, and generated images are available at https://github.com/valeoai/IGLOSS. |
|
| ConInfer: Context-Aware Inference for Training-Free Open-Vocabulary Remote Sensing Segmentation | 2026-03-31 | ShowTraining-free open-vocabulary remote sensing segmentation (OVRSS), empowered by vision-language models, has emerged as a promising paradigm for achieving category-agnostic semantic understanding in remote sensing imagery. Existing approaches mainly focus on enhancing feature representations or mitigating modality discrepancies to improve patch-level prediction accuracy. However, such independent prediction schemes are fundamentally misaligned with the intrinsic characteristics of remote sensing data. In real-world applications, remote sensing scenes are typically large-scale and exhibit strong spatial as well as semantic correlations, making isolated patch-wise predictions insufficient for accurate segmentation. To address this limitation, we propose ConInfer, a context-aware inference framework for OVRSS that performs joint prediction across multiple spatial units while explicitly modeling their inter-unit semantic dependencies. By incorporating global contextual cues, our method significantly enhances segmentation consistency, robustness, and generalization in complex remote sensing environments. Extensive experiments on multiple benchmark datasets demonstrate that our approach consistently surpasses state-of-the-art per-pixel VLM-based baselines such as SegEarth-OV, achieving average improvements of 2.80% and 6.13% on open-vocabulary semantic segmentation and object extraction tasks, respectively. The implementation code is available at: https://github.com/Dog-Yang/ConInfer |
|
| Looking Beyond the Window: Global-Local Aligned CLIP for Training-free Open-Vocabulary Semantic Segmentation | 2026-03-24 | ShowA sliding-window inference strategy is commonly adopted in recent training-free open-vocabulary semantic segmentation methods to overcome limitation of the CLIP in processing high-resolution images. However, this approach introduces a new challenge: each window is processed independently, leading to semantic discrepancy across windows. To address this issue, we propose Global-Local Aligned CLIP~(GLA-CLIP), a framework that facilitates comprehensive information exchange across windows. Rather than limiting attention to tokens within individual windows, GLA-CLIP extends key-value tokens to incorporate contextual cues from all windows. Nevertheless, we observe a window bias: outer-window tokens are less likely to be attended, since query features are produced through interactions within the inner window patches, thereby lacking semantic grounding beyond their local context. To mitigate this, we introduce a proxy anchor, constructed by aggregating tokens highly similar to the given query from all windows, which provides a unified semantic reference for measuring similarity across both inner- and outer-window patches. Furthermore, we propose a dynamic normalization scheme that adjusts attention strength according to object scale by dynamically scaling and thresholding the attention map to cope with small-object scenarios. Moreover, GLA-CLIP can be equipped on existing methods and broad their receptive field. Extensive experiments validate the effectiveness of GLA-CLIP in enhancing training-free open-vocabulary semantic segmentation performance. Code is available at https://github.com/2btlFe/GLA-CLIP. |
18 pa...18 pages, 13 figures, 12 tables, Accepted to CVPR 2026 |
| PEARL: Geometry Aligns Semantics for Training-Free Open-Vocabulary Semantic Segmentation | 2026-03-23 | ShowTraining-free open-vocabulary semantic segmentation (OVSS) promises rapid adaptation to new label sets without retraining. Yet, many methods rely on heavy post-processing or handle text and vision in isolation, leaving cross-modal geometry underutilized. Others introduce auxiliary vision backbones or multi-model pipelines, which increase complexity and latency while compromising design simplicity. We present PEARL, \textbf{\underline{P}}rocrust\textbf{\underline{e}}s \textbf{\underline{a}}lignment with text-awa\textbf{\underline{r}}e \textbf{\underline{L}}aplacian propagation, a compact two-step inference that follows an align-then-propagate principle. The Procrustes alignment step performs an orthogonal projection inside the last self-attention block, rotating keys toward the query subspace via a stable polar iteration. The text-aware Laplacian propagation then refines per-pixel logits on a small grid through a confidence-weighted, text-guided graph solve: text provides both a data-trust signal and neighbor gating, while image gradients preserve boundaries. In this work, our method is fully training-free, plug-and-play, and uses only fixed constants, adding minimal latency with a small per-head projection and a few conjugate-gradient steps. Our approach, PEARL, sets a new state-of-the-art in training-free OVSS without extra data or auxiliary backbones across standard benchmarks, achieving superior performance under both with-background and without-background protocols. |
accep...accepted by CVPR 2026 |
| dinov3.seg: Open-Vocabulary Semantic Segmentation with DINOv3 | 2026-03-19 | ShowOpen-Vocabulary Semantic Segmentation (OVSS) assigns pixel-level labels from an open set of text-defined categories, demanding reliable generalization to unseen classes at inference. Although modern vision-language models (VLMs) support strong open-vocabulary recognition, their representations learned through global contrastive objectives remain suboptimal for dense prediction, prompting many OVSS methods to depend on limited adaptation or refinement of image-text similarity maps. This, in turn, restricts spatial precision and robustness in complex, cluttered scenes. We introduce dinov3.seg, extending dinov3.txt into a dedicated framework for OVSS. Our contributions are four-fold. First, we design a task-specific architecture tailored to this backbone, systematically adapting established design principles from prior open-vocabulary segmentation work. Second, we jointly leverage text embeddings aligned with both the global [CLS] token and local patch-level visual features from ViT-based encoder, effectively combining semantic discrimination with fine-grained spatial locality. Third, unlike prior approaches that rely primarily on post hoc similarity refinement, we perform early refinement of visual representations prior to image-text interaction, followed by late refinement of the resulting image-text correlation features, enabling more accurate and robust dense predictions in cluttered scenes. Finally, we propose a high-resolution local-global inference strategy based on sliding-window aggregation, which preserves spatial detail while maintaining global context. We conduct extensive experiments on five widely adopted OVSS benchmarks to evaluate our approach. The results demonstrate its effectiveness and robustness, consistently outperforming current state-of-the-art methods. |
|
| DCP-CLIP:A Coarse-to-Fine Framework for Open-Vocabulary Semantic Segmentation with Dual Interaction | 2026-03-14 | ShowThe recent years have witnessed the remarkable development for open-vocabulary semantic segmentation (OVSS) using visual-language foundation models, yet still suffer from following fundamental challenges: (1) insufficient cross-modal communications between textual and visual spaces, and (2) significant computational costs from the interactions with massive number of categories. To address these issues, this paper describes a novel coarse-to-fine framework, called DCP-CLIP, for OVSS. Unlike prior efforts that mainly relied on pre-established category content and the inherent spatial-class interaction capability of CLIP, we dynamic constructing category-relevant textual features and explicitly models dual interactions between spatial image features and textual class semantics. Specifically, we first leverage CLIP's open-vocabulary recognition capability to identify semantic categories relevant to the image context, upon which we dynamically generate corresponding textual features to serve as initial textual guidance. Subsequently, we conduct a coarse segmentation by cross-modally integrating semantic information from textual guidance into the visual representations and achieve refined segmentation by integrating spatially enriched features from the encoder to recover fine-grained details and enhance spatial resolution. In final, we leverage spatial information from the segmentation side to refine category predictions for each mask, facilitating more precise semantic labeling. Experiments on multiple OVSS benchmarks demonstrate that DCP-CLIP outperforms existing methods by delivering both higher accuracy and greater efficiency. |
13 pages, 7 figures |
| JOPP-3D: Joint Open Vocabulary Semantic Segmentation on Point Clouds and Panoramas | 2026-03-12 | ShowSemantic segmentation across visual modalities such as 3D point clouds and panoramic images remains a challenging task, primarily due to the scarcity of annotated data and the limited adaptability of fixed-label models. In this paper, we present JOPP-3D, an open-vocabulary semantic segmentation framework that jointly leverages panoramic and point cloud data to enable language-driven scene understanding. We convert RGB-D panoramic images into their corresponding tangential perspective images and 3D point clouds, then use these modalities to extract and align foundational vision-language features. This allows natural language querying to generate semantic masks on both input modalities. Experimental evaluation on the Stanford-2D-3D-s and ToF-360 datasets demonstrates the capability of JOPP-3D to produce coherent and semantically meaningful segmentations across panoramic and 3D domains. Our proposed method achieves a significant improvement compared to the SOTA in open and closed vocabulary 2D and 3D semantic segmentation. |
|
| Open-Vocabulary Domain Generalization in Urban-Scene Segmentation | 2026-03-09 | ShowDomain Generalization in Semantic Segmentation (DG-SS) aims to enable segmentation models to perform robustly in unseen environments. However, conventional DG-SS methods are restricted to a fixed set of known categories, limiting their applicability in open-world scenarios. Recent progress in Vision-Language Models (VLMs) has advanced Open-Vocabulary Semantic Segmentation (OV-SS) by enabling models to recognize a broader range of concepts. Yet, these models remain sensitive to domain shifts and struggle to maintain robustness when deployed in unseen environments, a challenge that is particularly severe in urban-driving scenarios. To bridge this gap, we introduce Open-Vocabulary Domain Generalization in Semantic Segmentation (OVDG-SS), a new setting that jointly addresses unseen domains and unseen categories. We introduce the first benchmark for OVDG-SS in autonomous driving, addressing a previously unexplored problem and covering both synthetic-to-real and real-to-real generalization across diverse unseen domains and unseen categories. In OVDG-SS, we observe that domain shifts often distort text-image correlations in pre-trained VLMs, which hinders the performance of OV-SS models. To tackle this challenge, we propose S2-Corr, a state-space-driven text-image correlation refinement mechanism that mitigates domain-induced distortions and produces more consistent text-image correlations under distribution changes. Extensive experiments on our constructed benchmark demonstrate that the proposed method achieves superior cross-domain performance and efficiency compared to existing OV-SS approaches. |
|
| Open-Vocabulary Semantic Segmentation in Remote Sensing via Hierarchical Attention Masking and Model Composition | 2026-02-27 | ShowIn this paper, we propose ReSeg-CLIP, a new training-free Open-Vocabulary Semantic Segmentation method for remote sensing data. To compensate for the problems of vision language models, such as CLIP in semantic segmentation caused by inappropriate interactions within the self-attention layers, we introduce a hierarchical scheme utilizing masks generated by SAM to constrain the interactions at multiple scales. We also present a model composition approach that averages the parameters of multiple RS-specific CLIP variants, taking advantage of a new weighting scheme that evaluates representational quality using varying text prompts. Our method achieves state-of-the-art results across three RS benchmarks without additional training. |
Publi...Published in the proceedings of the British Machine Vision Conference Workshops 2025 |
| Enabling Training-Free Text-Based Remote Sensing Segmentation | 2026-02-19 | ShowRecent advances in Vision Language Models (VLMs) and Vision Foundation Models (VFMs) have opened new opportunities for zero-shot text-guided segmentation of remote sensing imagery. However, most existing approaches still rely on additional trainable components, limiting their generalisation and practical applicability. In this work, we investigate to what extent text-based remote sensing segmentation can be achieved without additional training, by relying solely on existing foundation models. We propose a simple yet effective approach that integrates contrastive and generative VLMs with the Segment Anything Model (SAM), enabling a fully training-free or lightweight LoRA-tuned pipeline. Our contrastive approach employs CLIP as mask selector for SAM's grid-based proposals, achieving state-of-the-art open-vocabulary semantic segmentation (OVSS) in a completely zero-shot setting. In parallel, our generative approach enables reasoning and referring segmentation by generating click prompts for SAM using GPT-5 in a zero-shot setting and a LoRA-tuned Qwen-VL model, with the latter yielding the best results. Extensive experiments across 19 remote sensing benchmarks, including open-vocabulary, referring, and reasoning-based tasks, demonstrate the strong capabilities of our approach. Code will be released at https://github.com/josesosajs/trainfree-rs-segmentation. |
|
| LMSeg: Unleashing the Power of Large-Scale Models for Open-Vocabulary Semantic Segmentation | 2026-02-18 | ShowIt is widely agreed that open-vocabulary-based approaches outperform classical closed-set training solutions for recognizing unseen objects in images for semantic segmentation. Existing open-vocabulary approaches leverage vision-language models, such as CLIP, to align visual features with rich semantic features acquired through pre-training on large-scale vision-language datasets. However, the text prompts employed in these methods are short phrases based on fixed templates, failing to capture comprehensive object attributes. Moreover, while the CLIP model excels at exploiting image-level features, it is less effective at pixel-level representation, which is crucial for semantic segmentation tasks. In this work, we propose to alleviate the above-mentioned issues by leveraging multiple large-scale models to enhance the alignment between fine-grained visual features and enriched linguistic features. Specifically, our method employs large language models (LLMs) to generate enriched language prompts with diverse visual attributes for each category, including color, shape/size, and texture/material. Additionally, for enhanced visual feature extraction, the SAM model is adopted as a supplement to the CLIP visual encoder through a proposed learnable weighted fusion strategy. Built upon these techniques, our method, termed LMSeg, achieves state-of-the-art performance across all major open-vocabulary segmentation benchmarks. The code will be made available soon. |
|
| LoGoSeg: Integrating Local and Global Features for Open-Vocabulary Semantic Segmentation | 2026-02-12 | ShowOpen-vocabulary semantic segmentation (OVSS) extends traditional closed-set segmentation by enabling pixel-wise annotation for both seen and unseen categories using arbitrary textual descriptions. While existing methods leverage vision-language models (VLMs) like CLIP, their reliance on image-level pretraining often results in imprecise spatial alignment, leading to mismatched segmentations in ambiguous or cluttered scenes. However, most existing approaches lack strong object priors and region-level constraints, which can lead to object hallucination or missed detections, further degrading performance. To address these challenges, we propose LoGoSeg, an efficient single-stage framework that integrates three key innovations: (i) an object existence prior that dynamically weights relevant categories through global image-text similarity, effectively reducing hallucinations; (ii) a region-aware alignment module that establishes precise region-level visual-textual correspondences; and (iii) a dual-stream fusion mechanism that optimally combines local structural information with global semantic context. Unlike prior works, LoGoSeg eliminates the need for external mask proposals, additional backbones, or extra datasets, ensuring efficiency. Extensive experiments on six benchmarks (A-847, PC-459, A-150, PC-59, PAS-20, and PAS-20b) demonstrate its competitive performance and strong generalization in open-vocabulary settings. |
|
| Geospatial-Reasoning-Driven Vocabulary-Agnostic Remote Sensing Semantic Segmentation | 2026-02-09 | ShowOpen-vocabulary semantic segmentation has emerged as a promising research direction in remote sensing, enabling the recognition of diverse land-cover types beyond pre-defined category sets. However, existing methods predominantly rely on the passive mapping of visual features and textual embeddings. This ``appearance-based" paradigm lacks geospatial contextual awareness, leading to severe semantic ambiguity and misclassification when encountering land-cover classes with similar spectral features but distinct semantic attributes. To address this, we propose a Geospatial Reasoning Chain-of-Thought (GR-CoT) framework designed to enhance the scene understanding capabilities of Multimodal Large Language Models (MLLMs), thereby guiding open-vocabulary segmentation models toward precise mapping. The framework comprises two collaborative components: an offline knowledge distillation stream and an online instance reasoning stream. The offline stream establishes fine-grained category interpretation standards to resolve semantic conflicts between similar land-cover types. During online inference, the framework executes a sequential reasoning process involving macro-scenario anchoring, visual feature decoupling, and knowledge-driven decision synthesis. This process generates an image-adaptive vocabulary that guides downstream models to achieve pixel-level alignment with correct geographical semantics. Extensive experiments on the LoveDA and GID5 benchmarks demonstrate the superiority of our approach. |
5 pages, 3 figures |
| DiSa: Saliency-Aware Foreground-Background Disentangled Framework for Open-Vocabulary Semantic Segmentation | 2026-01-27 | ShowOpen-vocabulary semantic segmentation aims to assign labels to every pixel in an image based on text labels. Existing approaches typically utilize vision-language models (VLMs), such as CLIP, for dense prediction. However, VLMs, pre-trained on image-text pairs, are biased toward salient, object-centric regions and exhibit two critical limitations when adapted to segmentation: (i) Foreground Bias, which tends to ignore background regions, and (ii) Limited Spatial Localization, resulting in blurred object boundaries. To address these limitations, we introduce DiSa, a novel saliency-aware foreground-background disentangled framework. By explicitly incorporating saliency cues in our designed Saliency-aware Disentanglement Module (SDM), DiSa separately models foreground and background ensemble features in a divide-and-conquer manner. Additionally, we propose a Hierarchical Refinement Module (HRM) that leverages pixel-wise spatial contexts and enables channel-wise feature refinement through multi-level updates. Extensive experiments on six benchmarks demonstrate that DiSa consistently outperforms state-of-the-art methods. |
19 pages, 11 figures |
| ARM: A Learnable, Plug-and-Play Module for CLIP-based Open-vocabulary Semantic Segmentation | 2025-12-30 | ShowOpen-vocabulary semantic segmentation (OVSS) is fundamentally hampered by the coarse, image-level representations of CLIP, which lack precise pixel-level details. Existing training-free methods attempt to resolve this by either importing priors from costly external foundation models (e.g., SAM, DINO) or by applying static, hand-crafted heuristics to CLIP's internal features. These approaches are either computationally expensive or sub-optimal. We propose the Attention Refinement Module (ARM), a lightweight, learnable module that effectively unlocks and refines CLIP's internal potential. Unlike static-fusion methods, ARM learns to adaptively fuse hierarchical features. It employs a semantically-guided cross-attention block, using robust deep features (K, V) to select and refine detail-rich shallow features (Q), followed by a self-attention block. The key innovation lies in a ``train once, use anywhere" paradigm. Trained once on a general-purpose dataset (e.g., COCO-Stuff), ARM acts as a universal plug-and-play post-processor for diverse training-free frameworks. Extensive experiments show that ARM consistently boosts baseline performance on multiple benchmarks with negligible inference overhead, establishing an efficient and effective paradigm for training-free OVSS. |
10 pages, 4 figures |
| Efficient Redundancy Reduction for Open-Vocabulary Semantic Segmentation | 2025-12-21 | ShowOpen-vocabulary semantic segmentation (OVSS) is an open-world task that aims to assign each pixel within an image to a specific class defined by arbitrary text descriptions. While large-scale vision-language models have shown remarkable open-vocabulary capabilities, their image-level pretraining limits effectiveness on pixel-wise dense prediction tasks like OVSS. Recent cost-based methods narrow this granularity gap by constructing pixel-text cost maps and refining them via cost aggregation mechanisms. Despite achieving promising performance, these approaches suffer from high computational costs and long inference latency. In this paper, we identify two major sources of redundancy in the cost-based OVSS framework: redundant information introduced during cost maps construction and inefficient sequence modeling in cost aggregation. To address these issues, we propose ERR-Seg, an efficient architecture that incorporates Redundancy-Reduced Hierarchical Cost maps (RRHC) and Redundancy-Reduced Cost Aggregation (RRCA). Specifically, RRHC reduces redundant class channels by customizing a compact class vocabulary for each image and integrates hierarchical cost maps to enrich semantic representation. RRCA alleviates computational burden by performing both spatial-level and class-level sequence reduction before aggregation. Overall, ERR-Seg results in a lightweight structure for OVSS, characterized by substantial memory and computational savings without compromising accuracy. Compared to previous state-of-the-art methods on the ADE20K-847 benchmark, ERR-Seg improves performance by |
|
| Leveraging Depth and Language for Open-Vocabulary Domain-Generalized Semantic Segmentation | 2025-12-11 | ShowOpen-Vocabulary semantic segmentation (OVSS) and domain generalization in semantic segmentation (DGSS) highlight a subtle complementarity that motivates Open-Vocabulary Domain-Generalized Semantic Segmentation (OV-DGSS). OV-DGSS aims to generate pixel-level masks for unseen categories while maintaining robustness across unseen domains, a critical capability for real-world scenarios such as autonomous driving in adverse conditions. We introduce Vireo, a novel single-stage framework for OV-DGSS that unifies the strengths of OVSS and DGSS for the first time. Vireo builds upon the frozen Visual Foundation Models (VFMs) and incorporates scene geometry via Depth VFMs to extract domain-invariant structural features. To bridge the gap between visual and textual modalities under domain shift, we propose three key components: (1) GeoText Prompts, which align geometric features with language cues and progressively refine VFM encoder representations; (2) Coarse Mask Prior Embedding (CMPE) for enhancing gradient flow for faster convergence and stronger textual influence; and (3) the Domain-Open-Vocabulary Vector Embedding Head (DOV-VEH), which fuses refined structural and semantic features for robust prediction. Comprehensive evaluation on these components demonstrates the effectiveness of our designs. Our proposed Vireo achieves the state-of-the-art performance and surpasses existing methods by a large margin in both domain generalization and open-vocabulary recognition, offering a unified and scalable solution for robust visual understanding in diverse and dynamic environments. Code is available at https://github.com/anonymouse-9c53tp182bvz/Vireo. |
Accep...Accepted by NeurIPS 2025 |
| Structure-Aware Feature Rectification with Region Adjacency Graphs for Training-Free Open-Vocabulary Semantic Segmentation | 2025-12-08 | ShowBenefiting from the inductive biases learned from large-scale datasets, open-vocabulary semantic segmentation (OVSS) leverages the power of vision-language models, such as CLIP, to achieve remarkable progress without requiring task-specific training. However, due to CLIP's pre-training nature on image-text pairs, it tends to focus on global semantic alignment, resulting in suboptimal performance when associating fine-grained visual regions with text. This leads to noisy and inconsistent predictions, particularly in local areas. We attribute this to a dispersed bias stemming from its contrastive training paradigm, which is difficult to alleviate using CLIP features alone. To address this, we propose a structure-aware feature rectification approach that incorporates instance-specific priors derived directly from the image. Specifically, we construct a region adjacency graph (RAG) based on low-level features (e.g., colour and texture) to capture local structural relationships and use it to refine CLIP features by enhancing local discrimination. Extensive experiments show that our method effectively suppresses segmentation noise, improves region-level consistency, and achieves strong performance on multiple open-vocabulary segmentation benchmarks. |
Accepted to WACV2026 |
| Open Vocabulary Compositional Explanations for Neuron Alignment | 2025-11-25 | ShowNeurons are the fundamental building blocks of deep neural networks, and their interconnections allow AI to achieve unprecedented results. Motivated by the goal of understanding how neurons encode information, compositional explanations leverage logical relationships between concepts to express the spatial alignment between neuron activations and human knowledge. However, these explanations rely on human-annotated datasets, restricting their applicability to specific domains and predefined concepts. This paper addresses this limitation by introducing a framework for the vision domain that allows users to probe neurons for arbitrary concepts and datasets. Specifically, the framework leverages masks generated by open vocabulary semantic segmentation to compute open vocabulary compositional explanations. The proposed framework consists of three steps: specifying arbitrary concepts, generating semantic segmentation masks using open vocabulary models, and deriving compositional explanations from these masks. The paper compares the proposed framework with previous methods for computing compositional explanations both in terms of quantitative metrics and human interpretability, analyzes the differences in explanations when shifting from human-annotated data to model-annotated data, and showcases the additional capabilities provided by the framework in terms of flexibility of the explanations with respect to the tasks and properties of interest. |
47 pages, 11 figures |
| SAM-MI: A Mask-Injected Framework for Enhancing Open-Vocabulary Semantic Segmentation with SAM | 2025-11-25 | ShowOpen-vocabulary semantic segmentation (OVSS) aims to segment and recognize objects universally. Trained on extensive high-quality segmentation data, the segment anything model (SAM) has demonstrated remarkable universal segmentation capabilities, offering valuable support for OVSS. Although previous methods have made progress in leveraging SAM for OVSS, there are still some challenges: (1) SAM's tendency to over-segment and (2) hard combinations between fixed masks and labels. This paper introduces a novel mask-injected framework, SAM-MI, which effectively integrates SAM with OVSS models to address these challenges. Initially, SAM-MI employs a Text-guided Sparse Point Prompter to sample sparse prompts for SAM instead of previous dense grid-like prompts, thus significantly accelerating the mask generation process. The framework then introduces Shallow Mask Aggregation (SMAgg) to merge partial masks to mitigate the SAM's over-segmentation issue. Finally, Decoupled Mask Injection (DMI) incorporates SAM-generated masks for guidance at low-frequency and high-frequency separately, rather than directly combining them with labels. Extensive experiments on multiple benchmarks validate the superiority of SAM-MI. Notably, the proposed method achieves a 16.7% relative improvement in mIoU over Grounded-SAM on the MESS benchmark, along with a 1.6$\times$ speedup. We hope SAM-MI can serve as an alternative methodology to effectively equip the OVSS model with SAM. |
|
| Target Refocusing via Attention Redistribution for Open-Vocabulary Semantic Segmentation: An Explainability Perspective | 2025-11-20 | ShowOpen-vocabulary semantic segmentation (OVSS) employs pixel-level vision-language alignment to associate category-related prompts with corresponding pixels. A key challenge is enhancing the multimodal dense prediction capability, specifically this pixel-level multimodal alignment. Although existing methods achieve promising results by leveraging CLIP's vision-language alignment, they rarely investigate the performance boundaries of CLIP for dense prediction from an interpretability mechanisms perspective. In this work, we systematically investigate CLIP's internal mechanisms and identify a critical phenomenon: analogous to human distraction, CLIP diverts significant attention resources from target regions to irrelevant tokens. Our analysis reveals that these tokens arise from dimension-specific over-activation; filtering them enhances CLIP's dense prediction performance. Consequently, we propose ReFocusing CLIP (RF-CLIP), a training-free approach that emulates human distraction-refocusing behavior to redirect attention from distraction tokens back to target regions, thereby refining CLIP's multimodal alignment granularity. Our method achieves SOTA performance on eight benchmarks while maintaining high inference efficiency. |
Accep...Accepted by AAAI 2026 |
| InfoCLIP: Bridging Vision-Language Pretraining and Open-Vocabulary Semantic Segmentation via Information-Theoretic Alignment Transfer | 2025-11-20 | ShowRecently, the strong generalization ability of CLIP has facilitated open-vocabulary semantic segmentation, which labels pixels using arbitrary text. However, existing methods that fine-tune CLIP for segmentation on limited seen categories often lead to overfitting and degrade the pretrained vision-language alignment. To stabilize modality alignment during fine-tuning, we propose InfoCLIP, which leverages an information-theoretic perspective to transfer alignment knowledge from pretrained CLIP to the segmentation task. Specifically, this transfer is guided by two novel objectives grounded in mutual information. First, we compress the pixel-text modality alignment from pretrained CLIP to reduce noise arising from its coarse-grained local semantic representations learned under image-text supervision. Second, we maximize the mutual information between the alignment knowledge of pretrained CLIP and the fine-tuned model to transfer compact local semantic relations suited for the segmentation task. Extensive evaluations across various benchmarks validate the effectiveness of InfoCLIP in enhancing CLIP fine-tuning for open-vocabulary semantic segmentation, demonstrating its adaptability and superiority in asymmetric transfer. |
Accep...Accepted by AAAI 2026 |
| FarSLIP: Discovering Effective CLIP Adaptation for Fine-Grained Remote Sensing Understanding | 2025-11-18 | ShowAs CLIP's global alignment limits its ability to capture fine-grained details, recent efforts have focused on enhancing its region-text alignment. However, current remote sensing (RS)-specific CLIP variants still inherit this limited spatial awareness. We identify two key limitations behind this: (1) current RS image-text datasets generate global captions from object-level labels, leaving the original object-level supervision underutilized; (2) despite the success of region-text alignment methods in general domain, their direct application to RS data often leads to performance degradation. To address these, we construct the first multi-granularity RS image-text dataset, MGRS-200k, featuring rich object-level textual supervision for RS region-category alignment. We further investigate existing fine-grained CLIP tuning strategies and find that current explicit region-text alignment methods, whether in a direct or indirect way, underperform due to severe degradation of CLIP's semantic coherence. Building on these, we propose FarSLIP, a Fine-grained Aligned RS Language-Image Pretraining framework. Rather than the commonly used patch-to-CLS self-distillation, FarSLIP employs patch-to-patch distillation to align local and global visual cues, which improves feature discriminability while preserving semantic coherence. Additionally, to effectively utilize region-text supervision, it employs simple CLS token-based region-category alignment rather than explicit patch-level alignment, further enhancing spatial awareness. FarSLIP features improved fine-grained vision-language alignment in RS domain and sets a new state of the art not only on RS open-vocabulary semantic segmentation, but also on image-level tasks such as zero-shot classification and image-text retrieval. Our dataset, code, and models are available at https://github.com/NJU-LHRS/FarSLIP. |
|
| SynSeg: Feature Synergy for Multi-Category Contrastive Learning in End-to-End Open-Vocabulary Semantic Segmentation | 2025-11-15 | ShowSemantic segmentation in open-vocabulary scenarios presents significant challenges due to the wide range and granularity of semantic categories. Existing weakly-supervised methods often rely on category-specific supervision and ill-suited feature construction methods for contrastive learning, leading to semantic misalignment and poor performance. In this work, we propose a novel weakly-supervised approach, SynSeg, to address the challenges. SynSeg performs Multi-Category Contrastive Learning (MCCL) as a stronger training signal with a new feature reconstruction framework named Feature Synergy Structure (FSS). Specifically, MCCL strategy robustly combines both intra- and inter-category alignment and separation in order to make the model learn the knowledge of correlations from different categories within the same image. Moreover, FSS reconstructs discriminative features for contrastive learning through prior fusion and semantic-activation-map enhancement, effectively avoiding the foreground bias introduced by the visual encoder. Furthermore, SynSeg is a lightweight end-to-end solution without using any mid-term output from large-scale pretrained models and capable for real-time inference. In general, SynSeg effectively improves the abilities in semantic localization and discrimination under weak supervision in an efficient manner. Extensive experiments on benchmarks demonstrate that our method outperforms state-of-the-art (SOTA) performance. Particularly, SynSeg achieves higher accuracy than SOTA baselines with a ratio from 6.9% up to 26.2%. |
|
| DGL-RSIS: Decoupling Global Spatial Context and Local Class Semantics for Training-Free Remote Sensing Image Segmentation | 2025-11-11 | ShowThe emergence of vision language models (VLMs) bridges the gap between vision and language, enabling multimodal understanding beyond traditional visual-only deep learning models. However, transferring VLMs from the natural image domain to remote sensing (RS) segmentation remains challenging due to the large domain gap and the diversity of RS inputs across tasks, particularly in open-vocabulary semantic segmentation (OVSS) and referring expression segmentation (RES). Here, we propose a training-free unified framework, termed DGL-RSIS, which decouples visual and textual representations and performs visual-language alignment at both local semantic and global contextual levels. Specifically, a Global-Local Decoupling (GLD) module decomposes textual inputs into local semantic tokens and global contextual tokens, while image inputs are partitioned into class-agnostic mask proposals. Then, a Local Visual-Textual Alignment (LVTA) module adaptively extracts context-aware visual features from the mask proposals and enriches textual features through knowledge-guided prompt engineering, achieving OVSS from a local perspective. Furthermore, a Global Visual-Textual Alignment (GVTA) module employs a global-enhanced Grad-CAM mechanism to capture contextual cues for referring expressions, followed by a mask selection module that integrates pixel-level activations into mask-level segmentation outputs, thereby achieving RES from a global perspective. Experiments on the iSAID (OVSS) and RRSIS-D (RES) benchmarks demonstrate that DGL-RSIS outperforms existing training-free approaches. Ablation studies further validate the effectiveness of each module. To the best of our knowledge, this is the first unified training-free framework for RS image segmentation, which effectively transfers the semantic capability of VLMs trained on natural images to the RS domain without additional training. |
|
| NERVE: Neighbourhood & Entropy-guided Random-walk for training free open-Vocabulary sEgmentation | 2025-11-11 | ShowDespite recent advances in Open-Vocabulary Semantic Segmentation (OVSS), existing training-free methods face several limitations: use of computationally expensive affinity refinement strategies, ineffective fusion of transformer attention maps due to equal weighting or reliance on fixed-size Gaussian kernels to reinforce local spatial smoothness, enforcing isotropic neighborhoods. We propose a strong baseline for training-free OVSS termed as NERVE (Neighbourhood & Entropy-guided Random-walk for open-Vocabulary sEgmentation), which uniquely integrates global and fine-grained local information, exploiting the neighbourhood structure from the self-attention layer of a stable diffusion model. We also introduce a stochastic random walk for refining the affinity rather than relying on fixed-size Gaussian kernels for local context. This spatial diffusion process encourages propagation across connected and semantically related areas, enabling it to effectively delineate objects with arbitrary shapes. Whereas most existing approaches treat self-attention maps from different transformer heads or layers equally, our method uses entropy-based uncertainty to select the most relevant maps. Notably, our method does not require any conventional post-processing techniques like Conditional Random Fields (CRF) or Pixel-Adaptive Mask Refinement (PAMR). Experiments are performed on 7 popular semantic segmentation benchmarks, yielding an overall state-of-the-art zero-shot segmentation performance, providing an effective approach to open-vocabulary semantic segmentation. |
|
| Test-Time Adaptation of Vision-Language Models for Open-Vocabulary Semantic Segmentation | 2025-11-09 | ShowRecently, test-time adaptation has attracted wide interest in the context of vision-language models for image classification. However, to the best of our knowledge, the problem is completely overlooked in dense prediction tasks such as Open-Vocabulary Semantic Segmentation (OVSS). In response, we propose a novel TTA method tailored to adapting VLMs for segmentation during test time. Unlike TTA methods for image classification, our Multi-Level and Multi-Prompt (MLMP) entropy minimization integrates features from intermediate vision-encoder layers and is performed with different text-prompt templates at both the global CLS token and local pixel-wise levels. Our approach could be used as plug-and-play for any segmentation network, does not require additional training data or labels, and remains effective even with a single test sample. Furthermore, we introduce a comprehensive OVSS TTA benchmark suite, which integrates a rigorous evaluation protocol, nine segmentation datasets, 15 common synthetic corruptions, and additional real and rendered domain shifts, \textbf{with a total of 87 distinct test scenarios}, establishing a standardized and comprehensive testbed for future TTA research in open-vocabulary segmentation. Our experiments on this suite demonstrate that our segmentation-tailored method consistently delivers significant gains over direct adoption of TTA classification baselines. Code and data are available at https://github.com/dosowiechi/MLMP. |
|
| RadZero: Similarity-Based Cross-Attention for Explainable Vision-Language Alignment in Chest X-ray with Zero-Shot Multi-Task Capability | 2025-11-06 | ShowRecent advancements in multimodal models have significantly improved vision-language (VL) alignment in radiology. However, existing approaches struggle to effectively utilize complex radiology reports for learning and offer limited interpretability through attention probability visualizations. To address these challenges, we introduce |
NeurIPS 2025 |
| Improving Visual Discriminability of CLIP for Training-Free Open-Vocabulary Semantic Segmentation | 2025-10-27 | ShowExtending CLIP models to semantic segmentation remains challenging due to the misalignment between their image-level pre-training objectives and the pixel-level visual understanding required for dense prediction. While prior efforts have achieved encouraging results by reorganizing the final layer and features, they often inherit the global alignment bias of preceding layers, leading to suboptimal segmentation performance. In this work, we propose LHT-CLIP, a novel training-free framework that systematically exploits the visual discriminability of CLIP across layer, head, and token levels. Through comprehensive analysis, we reveal three key insights: (i) the final layers primarily strengthen image-text alignment with sacrifice of visual discriminability (e.g., last 3 layers in ViT-B/16 and 8 layers in ViT-L/14), partly due to the emergence of anomalous tokens; (ii) a subset of attention heads (e.g., 10 out of 144 in ViT-B/16) display consistently strong visual discriminability across datasets; (iii) abnormal tokens display sparse and consistent activation pattern compared to normal tokens. Based on these findings, we propose three complementary techniques: semantic-spatial reweighting, selective head enhancement, and abnormal token replacement to effectively restore visual discriminability and improve segmentation performance without any additional training, auxiliary pre-trained networks, or extensive hyperparameter tuning. Extensive experiments on 8 common semantic segmentation benchmarks demonstrate that LHT-CLIP achieves state-of-the-art performance across diverse scenarios, highlighting its effectiveness and practicality for real-world deployment. |
23 pa...23 pages, 10 figures, 14 tables |
| Neuro-Symbolic Spatial Reasoning in Segmentation | 2025-10-17 | ShowOpen-Vocabulary Semantic Segmentation (OVSS) assigns pixel-level labels from an open set of categories, requiring generalization to unseen and unlabelled objects. Using vision-language models (VLMs) to correlate local image patches with potential unseen object categories suffers from a lack of understanding of spatial relations of objects in a scene. To solve this problem, we introduce neuro-symbolic (NeSy) spatial reasoning in OVSS. In contrast to contemporary VLM correlation-based approaches, we propose Relational Segmentor (RelateSeg) to impose explicit spatial relational constraints by first order logic (FOL) formulated in a neural network architecture. This is the first attempt to explore NeSy spatial reasoning in OVSS. Specifically, RelateSeg automatically extracts spatial relations, e.g., <cat, to-right-of, person>, and encodes them as first-order logic formulas using our proposed pseudo categories. Each pixel learns to predict both a semantic category (e.g., "cat") and a spatial pseudo category (e.g., "right of person") simultaneously, enforcing relational constraints (e.g., a "cat" pixel must lie to the right of a "person"). Finally, these logic constraints are formulated in a deep network architecture by fuzzy logic relaxation, enabling end-to-end learning of spatial-relationally consistent segmentation. RelateSeg achieves state-of-the-art performance in terms of average mIoU across four benchmark datasets and particularly shows clear advantages on images containing multiple categories, with the cost of only introducing a single auxiliary loss function and no additional parameters, validating the effectiveness of NeSy spatial reasoning in OVSS. |
|
| Lost in Translation? Vocabulary Alignment for Source-Free Adaptation in Open-Vocabulary Semantic Segmentation | 2025-09-29 | ShowWe introduce VocAlign, a novel source-free domain adaptation framework specifically designed for VLMs in open-vocabulary semantic segmentation. Our method adopts a student-teacher paradigm enhanced with a vocabulary alignment strategy, which improves pseudo-label generation by incorporating additional class concepts. To ensure efficiency, we use Low-Rank Adaptation (LoRA) to fine-tune the model, preserving its original capabilities while minimizing computational overhead. In addition, we propose a Top-K class selection mechanism for the student model, which significantly reduces memory requirements while further improving adaptation performance. Our approach achieves a notable 6.11 mIoU improvement on the CityScapes dataset and demonstrates superior performance on zero-shot segmentation benchmarks, setting a new standard for source-free adaptation in the open-vocabulary setting. |
BMVC ...BMVC 2025 - Project Page: https://thegoodailab.org/blog/vocalign - Code: https://github.com/Sisso16/VocAlign |
| OpenUrban3D: Annotation-Free Open-Vocabulary Semantic Segmentation of Large-Scale Urban Point Clouds | 2025-09-13 | ShowOpen-vocabulary semantic segmentation enables models to recognize and segment objects from arbitrary natural language descriptions, offering the flexibility to handle novel, fine-grained, or functionally defined categories beyond fixed label sets. While this capability is crucial for large-scale urban point clouds that support applications such as digital twins, smart city management, and urban analytics, it remains largely unexplored in this domain. The main obstacles are the frequent absence of high-quality, well-aligned multi-view imagery in large-scale urban point cloud datasets and the poor generalization of existing three-dimensional (3D) segmentation pipelines across diverse urban environments with substantial variation in geometry, scale, and appearance. To address these challenges, we present OpenUrban3D, the first 3D open-vocabulary semantic segmentation framework for large-scale urban scenes that operates without aligned multi-view images, pre-trained point cloud segmentation networks, or manual annotations. Our approach generates robust semantic features directly from raw point clouds through multi-view, multi-granularity rendering, mask-level vision-language feature extraction, and sample-balanced fusion, followed by distillation into a 3D backbone model. This design enables zero-shot segmentation for arbitrary text queries while capturing both semantic richness and geometric priors. Extensive experiments on large-scale urban benchmarks, including SensatUrban and SUM, show that OpenUrban3D achieves significant improvements in both segmentation accuracy and cross-scene generalization over existing methods, demonstrating its potential as a flexible and scalable solution for 3D urban scene understanding. |
|
| Novel Category Discovery with X-Agent Attention for Open-Vocabulary Semantic Segmentation | 2025-09-03 | ShowOpen-vocabulary semantic segmentation (OVSS) conducts pixel-level classification via text-driven alignment, where the domain discrepancy between base category training and open-vocabulary inference poses challenges in discriminative modeling of latent unseen category. To address this challenge, existing vision-language model (VLM)-based approaches demonstrate commendable performance through pre-trained multi-modal representations. However, the fundamental mechanisms of latent semantic comprehension remain underexplored, making the bottleneck for OVSS. In this work, we initiate a probing experiment to explore distribution patterns and dynamics of latent semantics in VLMs under inductive learning paradigms. Building on these insights, we propose X-Agent, an innovative OVSS framework employing latent semantic-aware ``agent'' to orchestrate cross-modal attention mechanisms, simultaneously optimizing latent semantic dynamic and amplifying its perceptibility. Extensive benchmark evaluations demonstrate that X-Agent achieves state-of-the-art performance while effectively enhancing the latent semantic saliency. |
Accep...Accepted by ACMMM2025 |
| Annotation-Free Open-Vocabulary Segmentation for Remote-Sensing Images | 2025-08-25 | ShowSemantic segmentation of remote sensing (RS) images is pivotal for comprehensive Earth observation, but the demand for interpreting new object categories, coupled with the high expense of manual annotation, poses significant challenges. Although open-vocabulary semantic segmentation (OVSS) offers a promising solution, existing frameworks designed for natural images are insufficient for the unique complexities of RS data. They struggle with vast scale variations and fine-grained details, and their adaptation often relies on extensive, costly annotations. To address this critical gap, this paper introduces SegEarth-OV, the first framework for annotation-free open-vocabulary segmentation of RS images. Specifically, we propose SimFeatUp, a universal upsampler that robustly restores high-resolution spatial details from coarse features, correcting distorted target shapes without any task-specific post-training. We also present a simple yet effective Global Bias Alleviation operation to subtract the inherent global context from patch features, significantly enhancing local semantic fidelity. These components empower SegEarth-OV to effectively harness the rich semantics of pre-trained VLMs, making OVSS possible in optical RS contexts. Furthermore, to extend the framework's universality to other challenging RS modalities like SAR images, where large-scale VLMs are unavailable and expensive to create, we introduce AlignEarth, which is a distillation-based strategy and can efficiently transfer semantic knowledge from an optical VLM encoder to an SAR encoder, bypassing the need to build SAR foundation models from scratch and enabling universal OVSS across diverse sensor types. Extensive experiments on both optical and SAR datasets validate that SegEarth-OV can achieve dramatic improvements over the SOTA methods, establishing a robust foundation for annotation-free and open-world Earth observation. |
All c...All codes and models will be released at https://github.com/earth-insights/SegEarth-OV-2 |
| CitySeg: A 3D Open Vocabulary Semantic Segmentation Foundation Model in City-scale Scenarios | 2025-08-13 | ShowSemantic segmentation of city-scale point clouds is a critical technology for Unmanned Aerial Vehicle (UAV) perception systems, enabling the classification of 3D points without relying on any visual information to achieve comprehensive 3D understanding. However, existing models are frequently constrained by the limited scale of 3D data and the domain gap between datasets, which lead to reduced generalization capability. To address these challenges, we propose CitySeg, a foundation model for city-scale point cloud semantic segmentation that incorporates text modality to achieve open vocabulary segmentation and zero-shot inference. Specifically, in order to mitigate the issue of non-uniform data distribution across multiple domains, we customize the data preprocessing rules, and propose a local-global cross-attention network to enhance the perception capabilities of point networks in UAV scenarios. To resolve semantic label discrepancies across datasets, we introduce a hierarchical classification strategy. A hierarchical graph established according to the data annotation rules consolidates the data labels, and the graph encoder is used to model the hierarchical relationships between categories. In addition, we propose a two-stage training strategy and employ hinge loss to increase the feature separability of subcategories. Experimental results demonstrate that the proposed CitySeg achieves state-of-the-art (SOTA) performance on nine closed-set benchmarks, significantly outperforming existing approaches. Moreover, for the first time, CitySeg enables zero-shot generalization in city-scale point cloud scenarios without relying on visual information. |
|
| Training-Free Class Purification for Open-Vocabulary Semantic Segmentation | 2025-08-01 | ShowFine-tuning pre-trained vision-language models has emerged as a powerful approach for enhancing open-vocabulary semantic segmentation (OVSS). However, the substantial computational and resource demands associated with training on large datasets have prompted interest in training-free methods for OVSS. Existing training-free approaches primarily focus on modifying model architectures and generating prototypes to improve segmentation performance. However, they often neglect the challenges posed by class redundancy, where multiple categories are not present in the current test image, and visual-language ambiguity, where semantic similarities among categories create confusion in class activation. These issues can lead to suboptimal class activation maps and affinity-refined activation maps. Motivated by these observations, we propose FreeCP, a novel training-free class purification framework designed to address these challenges. FreeCP focuses on purifying semantic categories and rectifying errors caused by redundancy and ambiguity. The purified class representations are then leveraged to produce final segmentation predictions. We conduct extensive experiments across eight benchmarks to validate FreeCP's effectiveness. Results demonstrate that FreeCP, as a plug-and-play module, significantly boosts segmentation performance when combined with other OVSS methods. |
Accep...Accepted to ICCV 2025 |
| OpenSeg-R: Improving Open-Vocabulary Segmentation via Step-by-Step Visual Reasoning | 2025-08-01 | ShowOpen-Vocabulary Segmentation (OVS) has drawn increasing attention for its capacity to generalize segmentation beyond predefined categories. However, existing methods typically predict segmentation masks with simple forward inference, lacking explicit reasoning and interpretability. This makes it challenging for OVS model to distinguish similar categories in open-world settings due to the lack of contextual understanding and discriminative visual cues. To address this limitation, we propose a step-by-step visual reasoning framework for open-vocabulary segmentation, named OpenSeg-R. The proposed OpenSeg-R leverages Large Multimodal Models (LMMs) to perform hierarchical visual reasoning before segmentation. Specifically, we generate both generic and image-specific reasoning for each image, forming structured triplets that explain the visual reason for objects in a coarse-to-fine manner. Based on these reasoning steps, we can compose detailed description prompts, and feed them to the segmentor to produce more accurate segmentation masks. To the best of our knowledge, OpenSeg-R is the first framework to introduce explicit step-by-step visual reasoning into OVS. Experimental results demonstrate that OpenSeg-R significantly outperforms state-of-the-art methods on open-vocabulary semantic segmentation across five benchmark datasets. Moreover, it achieves consistent gains across all metrics on open-vocabulary panoptic segmentation. Qualitative results further highlight the effectiveness of our reasoning-guided framework in improving both segmentation precision and interpretability. Our code is publicly available at https://github.com/Hanzy1996/OpenSeg-R. |
|
| CorrCLIP: Reconstructing Patch Correlations in CLIP for Open-Vocabulary Semantic Segmentation | 2025-08-01 | ShowOpen-vocabulary semantic segmentation aims to assign semantic labels to each pixel without being constrained by a predefined set of categories. While Contrastive Language-Image Pre-training (CLIP) excels in zero-shot classification, it struggles to align image patches with category embeddings because of its incoherent patch correlations. This study reveals that inter-class correlations are the main reason for impairing CLIP's segmentation performance. Accordingly, we propose CorrCLIP, which reconstructs the scope and value of patch correlations. Specifically, CorrCLIP leverages the Segment Anything Model (SAM) to define the scope of patch interactions, reducing inter-class correlations. To mitigate the problem that SAM-generated masks may contain patches belonging to different classes, CorrCLIP incorporates self-supervised models to compute coherent similarity values, suppressing the weight of inter-class correlations. Additionally, we introduce two additional branches to strengthen patch features' spatial details and semantic representation. Finally, we update segmentation maps with SAM-generated masks to improve spatial consistency. Based on the improvement across patch correlations, feature representations, and segmentation maps, CorrCLIP achieves superior performance across eight benchmarks. Codes are available at: https://github.com/zdk258/CorrCLIP. |
Accep...Accepted to ICCV 2025 Oral |
| FLOSS: Free Lunch in Open-vocabulary Semantic Segmentation | 2025-07-30 | ShowIn this paper, we challenge the conventional practice in Open-Vocabulary Semantic Segmentation (OVSS) of using averaged class-wise text embeddings, which are typically obtained by encoding each class name with multiple templates (e.g., a photo of , a sketch of a ). We investigate the impact of templates for OVSS, and find that for each class, there exist single-template classifiers--which we refer to as class-experts--that significantly outperform the conventional averaged classifier. First, to identify these class-experts, we introduce a novel approach that estimates them without any labeled data or training. By leveraging the class-wise prediction entropy of single-template classifiers, we select those yielding the lowest entropy as the most reliable class-experts. Second, we combine the outputs of class-experts in a new fusion process. Our plug-and-play method, coined FLOSS, is orthogonal and complementary to existing OVSS methods, offering an improvement without the need for additional labels or training. Extensive experiments show that FLOSS consistently enhances state-of-the-art OVSS models, generalizes well across datasets with different distribution shifts, and delivers substantial improvements in low-data scenarios where only a few unlabeled images are available. Our code is available at https://github.com/yasserben/FLOSS . |
ICCV ...ICCV 2025; Project Page: https://yasserben.github.io/FLOSS/ |
| Personalized OVSS: Understanding Personal Concept in Open-Vocabulary Semantic Segmentation | 2025-07-15 | ShowWhile open-vocabulary semantic segmentation (OVSS) can segment an image into semantic regions based on arbitrarily given text descriptions even for classes unseen during training, it fails to understand personal texts (e.g., |
Accep...Accepted to ICCV 2025; 15 pages |
| Beyond-Labels: Advancing Open-Vocabulary Segmentation With Vision-Language Models | 2025-07-02 | ShowOpen-vocabulary semantic segmentation attempts to classify and outline objects in an image using arbitrary text labels, including those unseen during training. Self-supervised learning resolves numerous visual and linguistic processing problems when effectively trained. This study investigates simple yet efficient methods for adapting previously learned foundation models for open-vocabulary semantic segmentation tasks. Our research proposes "Beyond-Labels", a lightweight transformer-based fusion module that uses a small amount of image segmentation data to fuse frozen visual representations with language concepts. This strategy allows the model to leverage the extensive knowledge of pre-trained models without requiring significant retraining, making the approach data-efficient and scalable. Furthermore, we capture positional information in images using Fourier embeddings, improving generalization and enabling smooth and consistent spatial encoding. We perform thorough ablation studies to examine the main components of our proposed method. On the standard benchmark PASCAL-5i, the method performs better despite being trained on frozen vision and language representations. Index Terms: Beyond-Labels, open-vocabulary semantic segmentation, Fourier embeddings, PASCAL-5i |
Accep...Accepted at the 17th IEEE International Conference on Advanced Computational Intelligence (ICACI 2025) |
| Diffusion-Based Image Augmentation for Semantic Segmentation in Outdoor Robotics | 2025-06-30 | ShowThe performance of leaning-based perception algorithms suffer when deployed in out-of-distribution and underrepresented environments. Outdoor robots are particularly susceptible to rapid changes in visual scene appearance due to dynamic lighting, seasonality and weather effects that lead to scenes underrepresented in the training data of the learning-based perception system. In this conceptual paper, we focus on preparing our autonomous vehicle for deployment in snow-filled environments. We propose a novel method for diffusion-based image augmentation to more closely represent the deployment environment in our training data. Diffusion-based image augmentations rely on the public availability of vision foundation models learned on internet-scale datasets. The diffusion-based image augmentations allow us to take control over the semantic distribution of the ground surfaces in the training data and to fine-tune our model for its deployment environment. We employ open vocabulary semantic segmentation models to filter out augmentation candidates that contain hallucinations. We believe that diffusion-based image augmentations can be extended to many other environments apart from snow surfaces, like sandy environments and volcanic terrains. |
Prese...Presented at the 2025 IEEE ICRA Workshop on Field Robotics |
| Unleashing the Multi-View Fusion Potential: Noise Correction in VLM for Open-Vocabulary 3D Scene Understanding | 2025-06-28 | ShowRecent open-vocabulary 3D scene understanding approaches mainly focus on training 3D networks through contrastive learning with point-text pairs or by distilling 2D features into 3D models via point-pixel alignment. While these methods show considerable performance in benchmarks with limited vocabularies, they struggle to handle diverse object categories as the limited amount of 3D data upbound training strong open-vocabulary 3d models. We observe that 2D multi-view fusion methods take precedence in understanding diverse concepts in 3D scenes. However, inherent noises in vision-language models lead multi-view fusion to sub-optimal performance. To this end, we introduce MVOV3D, a novel approach aimed at unleashing the potential of 2D multi-view fusion for open-vocabulary 3D scene understanding. We focus on reducing the inherent noises without training, thereby preserving the generalizability while enhancing open-world capabilities. Specifically, MVOV3D improves multi-view 2D features by leveraging precise region-level image features and text features encoded by CLIP encoders and incorporates 3D geometric priors to optimize multi-view fusion. Extensive experiments on various datasets demonstrate the effectiveness of our method. Notably, our MVOV3D achieves a new record with 14.7% mIoU on ScanNet200 and 16.2% mIoU on Matterport160 for challenge open-vocabulary semantic segmentation, outperforming current leading trained 3D networks by a significant margin. |
|
| ReME: A Data-Centric Framework for Training-Free Open-Vocabulary Segmentation | 2025-06-27 | ShowTraining-free open-vocabulary semantic segmentation (OVS) aims to segment images given a set of arbitrary textual categories without costly model fine-tuning. Existing solutions often explore attention mechanisms of pre-trained models, such as CLIP, or generate synthetic data and design complex retrieval processes to perform OVS. However, their performance is limited by the capability of reliant models or the suboptimal quality of reference sets. In this work, we investigate the largely overlooked data quality problem for this challenging dense scene understanding task, and identify that a high-quality reference set can significantly benefit training-free OVS. With this observation, we introduce a data-quality-oriented framework, comprising a data pipeline to construct a reference set with well-paired segment-text embeddings and a simple similarity-based retrieval to unveil the essential effect of data. Remarkably, extensive evaluations on ten benchmark datasets demonstrate that our method outperforms all existing training-free OVS approaches, highlighting the importance of data-centric design for advancing OVS without training. Our code is available at https://github.com/xiweix/ReME . |
Accep...Accepted to ICCV 2025 |
| LPOSS: Label Propagation Over Patches and Pixels for Open-vocabulary Semantic Segmentation | 2025-06-25 | ShowWe propose a training-free method for open-vocabulary semantic segmentation using Vision-and-Language Models (VLMs). Our approach enhances the initial per-patch predictions of VLMs through label propagation, which jointly optimizes predictions by incorporating patch-to-patch relationships. Since VLMs are primarily optimized for cross-modal alignment and not for intra-modal similarity, we use a Vision Model (VM) that is observed to better capture these relationships. We address resolution limitations inherent to patch-based encoders by applying label propagation at the pixel level as a refinement step, significantly improving segmentation accuracy near class boundaries. Our method, called LPOSS+, performs inference over the entire image, avoiding window-based processing and thereby capturing contextual interactions across the full image. LPOSS+ achieves state-of-the-art performance among training-free methods, across a diverse set of datasets. Code: https://github.com/vladan-stojnic/LPOSS |
|
| Test-time Contrastive Concepts for Open-world Semantic Segmentation with Vision-Language Models | 2025-06-16 | ShowRecent CLIP-like Vision-Language Models (VLMs), pre-trained on large amounts of image-text pairs to align both modalities with a simple contrastive objective, have paved the way to open-vocabulary semantic segmentation. Given an arbitrary set of textual queries, image pixels are assigned the closest query in feature space. However, this works well when a user exhaustively lists all possible visual concepts in an image that contrast against each other for the assignment. This corresponds to the current evaluation setup in the literature, which relies on having access to a list of in-domain relevant concepts, typically classes of a benchmark dataset. Here, we consider the more challenging (and realistic) scenario of segmenting a single concept, given a textual prompt and nothing else. To achieve good results, besides contrasting with the generic 'background' text, we propose two different approaches to automatically generate, at test time, query-specific textual contrastive concepts. We do so by leveraging the distribution of text in the VLM's training set or crafted LLM prompts. We also propose a metric designed to evaluate this scenario and show the relevance of our approach on commonly used datasets. |
TMLR camera-ready |
| AetherVision-Bench: An Open-Vocabulary RGB-Infrared Benchmark for Multi-Angle Segmentation across Aerial and Ground Perspectives | 2025-06-04 | ShowOpen-vocabulary semantic segmentation (OVSS) involves assigning labels to each pixel in an image based on textual descriptions, leveraging world models like CLIP. However, they encounter significant challenges in cross-domain generalization, hindering their practical efficacy in real-world applications. Embodied AI systems are transforming autonomous navigation for ground vehicles and drones by enhancing their perception abilities, and in this study, we present AetherVision-Bench, a benchmark for multi-angle segmentation across aerial, and ground perspectives, which facilitates an extensive evaluation of performance across different viewing angles and sensor modalities. We assess state-of-the-art OVSS models on the proposed benchmark and investigate the key factors that impact the performance of zero-shot transfer models. Our work pioneers the creation of a robustness benchmark, offering valuable insights and establishing a foundation for future research. |
Accep...Accepted at Workshop on Foundation Models Meet Embodied Agents at CVPR 2025 (Non-archival Track) |
| OV-COAST: Cost Aggregation with Optimal Transport for Open-Vocabulary Semantic Segmentation | 2025-06-04 | ShowOpen-vocabulary semantic segmentation (OVSS) entails assigning semantic labels to each pixel in an image using textual descriptions, typically leveraging world models such as CLIP. To enhance out-of-domain generalization, we propose Cost Aggregation with Optimal Transport (OV-COAST) for open-vocabulary semantic segmentation. To align visual-language features within the framework of optimal transport theory, we employ cost volume to construct a cost matrix, which quantifies the distance between two distributions. Our approach adopts a two-stage optimization strategy: in the first stage, the optimal transport problem is solved using cost volume via Sinkhorn distance to obtain an alignment solution; in the second stage, this solution is used to guide the training of the CAT-Seg model. We evaluate state-of-the-art OVSS models on the MESS benchmark, where our approach notably improves the performance of the cost-aggregation model CAT-Seg with ViT-B backbone, achieving superior results, surpassing CAT-Seg by 1.72 % and SAN-B by 4.9 % mIoU. The code is available at https://github.com/adityagandhamal/OV-COAST/}{https://github.com/adityagandhamal/OV-COAST/ . |
Accep...Accepted at CVPR 2025 Workshop on Transformers for Vision (Non-archival track) |
| A Survey on Training-free Open-Vocabulary Semantic Segmentation | 2025-05-28 | ShowSemantic segmentation is one of the most fundamental tasks in image understanding with a long history of research, and subsequently a myriad of different approaches. Traditional methods strive to train models up from scratch, requiring vast amounts of computational resources and training data. In the advent of moving to open-vocabulary semantic segmentation, which asks models to classify beyond learned categories, large quantities of finely annotated data would be prohibitively expensive. Researchers have instead turned to training-free methods where they leverage existing models made for tasks where data is more easily acquired. Specifically, this survey will cover the history, nuance, idea development and the state-of-the-art in training-free open-vocabulary semantic segmentation that leverages existing multi-modal classification models. We will first give a preliminary on the task definition followed by an overview of popular model archetypes and then spotlight over 30 approaches split into broader research branches: purely CLIP-based, those leveraging auxiliary visual foundation models and ones relying on generative methods. Subsequently, we will discuss the limitations and potential problems of current research, as well as provide some underexplored ideas for future study. We believe this survey will serve as a good onboarding read to new researchers and spark increased interest in the area. |
|
| DPSeg: Dual-Prompt Cost Volume Learning for Open-Vocabulary Semantic Segmentation | 2025-05-16 | ShowOpen-vocabulary semantic segmentation aims to segment images into distinct semantic regions for both seen and unseen categories at the pixel level. Current methods utilize text embeddings from pre-trained vision-language models like CLIP but struggle with the inherent domain gap between image and text embeddings, even after extensive alignment during training. Additionally, relying solely on deep text-aligned features limits shallow-level feature guidance, which is crucial for detecting small objects and fine details, ultimately reducing segmentation accuracy. To address these limitations, we propose a dual prompting framework, DPSeg, for this task. Our approach combines dual-prompt cost volume generation, a cost volume-guided decoder, and a semantic-guided prompt refinement strategy that leverages our dual prompting scheme to mitigate alignment issues in visual prompt generation. By incorporating visual embeddings from a visual prompt encoder, our approach reduces the domain gap between text and image embeddings while providing multi-level guidance through shallow features. Extensive experiments demonstrate that our method significantly outperforms existing state-of-the-art approaches on multiple public datasets. |
Accepted by CVPR2025 |
| 3D Vision-Language Gaussian Splatting | 2025-05-05 | ShowRecent advancements in 3D reconstruction methods and vision-language models have propelled the development of multi-modal 3D scene understanding, which has vital applications in robotics, autonomous driving, and virtual/augmented reality. However, current multi-modal scene understanding approaches have naively embedded semantic representations into 3D reconstruction methods without striking a balance between visual and language modalities, which leads to unsatisfying semantic rasterization of translucent or reflective objects, as well as over-fitting on color modality. To alleviate these limitations, we propose a solution that adequately handles the distinct visual and semantic modalities, i.e., a 3D vision-language Gaussian splatting model for scene understanding, to put emphasis on the representation learning of language modality. We propose a novel cross-modal rasterizer, using modality fusion along with a smoothed semantic indicator for enhancing semantic rasterization. We also employ a camera-view blending technique to improve semantic consistency between existing and synthesized views, thereby effectively mitigating over-fitting. Extensive experiments demonstrate that our method achieves state-of-the-art performance in open-vocabulary semantic segmentation, surpassing existing methods by a significant margin. |
Accep...Accepted at ICLR 2025. Main paper + supplementary material |
| ITACLIP: Boosting Training-Free Semantic Segmentation with Image, Text, and Architectural Enhancements | 2025-04-14 | ShowRecent advances in foundational Vision Language Models (VLMs) have reshaped the evaluation paradigm in computer vision tasks. These foundational models, especially CLIP, have accelerated research in open-vocabulary computer vision tasks, including Open-Vocabulary Semantic Segmentation (OVSS). Although the initial results are promising, the dense prediction capabilities of VLMs still require further improvement. In this study, we enhance the semantic segmentation performance of CLIP by introducing new modules and modifications: 1) architectural changes in the last layer of ViT and the incorporation of attention maps from the middle layers with the last layer, 2) Image Engineering: applying data augmentations to enrich input image representations, and 3) using Large Language Models (LLMs) to generate definitions and synonyms for each class name to leverage CLIP's open-vocabulary capabilities. Our training-free method, ITACLIP, outperforms current state-of-the-art approaches on segmentation benchmarks such as COCO-Stuff, COCO-Object, Pascal Context, and Pascal VOC. Our code is available at https://github.com/m-arda-aydn/ITACLIP. |
|
| econSG: Efficient and Multi-view Consistent Open-Vocabulary 3D Semantic Gaussians | 2025-04-08 | ShowThe primary focus of most recent works on open-vocabulary neural fields is extracting precise semantic features from the VLMs and then consolidating them efficiently into a multi-view consistent 3D neural fields representation. However, most existing works over-trusted SAM to regularize image-level CLIP without any further refinement. Moreover, several existing works improved efficiency by dimensionality reduction of semantic features from 2D VLMs before fusing with 3DGS semantic fields, which inevitably leads to multi-view inconsistency. In this work, we propose econSG for open-vocabulary semantic segmentation with 3DGS. Our econSG consists of: 1) A Confidence-region Guided Regularization (CRR) that mutually refines SAM and CLIP to get the best of both worlds for precise semantic features with complete and precise boundaries. 2) A low dimensional contextual space to enforce 3D multi-view consistency while improving computational efficiency by fusing backprojected multi-view 2D features and follow by dimensional reduction directly on the fused 3D features instead of operating on each 2D view separately. Our econSG shows state-of-the-art performance on four benchmark datasets compared to the existing methods. Furthermore, we are also the most efficient training among all the methods. |
|
| Distilling Spectral Graph for Object-Context Aware Open-Vocabulary Semantic Segmentation | 2025-04-07 | ShowOpen-Vocabulary Semantic Segmentation (OVSS) has advanced with recent vision-language models (VLMs), enabling segmentation beyond predefined categories through various learning schemes. Notably, training-free methods offer scalable, easily deployable solutions for handling unseen data, a key goal of OVSS. Yet, a critical issue persists: lack of object-level context consideration when segmenting complex objects in the challenging environment of OVSS based on arbitrary query prompts. This oversight limits models' ability to group semantically consistent elements within object and map them precisely to user-defined arbitrary classes. In this work, we introduce a novel approach that overcomes this limitation by incorporating object-level contextual knowledge within images. Specifically, our model enhances intra-object consistency by distilling spectral-driven features from vision foundation models into the attention mechanism of the visual encoder, enabling semantically coherent components to form a single object mask. Additionally, we refine the text embeddings with zero-shot object presence likelihood to ensure accurate alignment with the specific objects represented in the images. By leveraging object-level contextual knowledge, our proposed approach achieves state-of-the-art performance with strong generalizability across diverse datasets. |
|
| Effective SAM Combination for Open-Vocabulary Semantic Segmentation | 2025-03-30 | ShowOpen-vocabulary semantic segmentation aims to assign pixel-level labels to images across an unlimited range of classes. Traditional methods address this by sequentially connecting a powerful mask proposal generator, such as the Segment Anything Model (SAM), with a pre-trained vision-language model like CLIP. But these two-stage approaches often suffer from high computational costs, memory inefficiencies. In this paper, we propose ESC-Net, a novel one-stage open-vocabulary segmentation model that leverages the SAM decoder blocks for class-agnostic segmentation within an efficient inference framework. By embedding pseudo prompts generated from image-text correlations into SAM's promptable segmentation framework, ESC-Net achieves refined spatial aggregation for accurate mask predictions. ESC-Net achieves superior performance on standard benchmarks, including ADE20K, PASCAL-VOC, and PASCAL-Context, outperforming prior methods in both efficiency and accuracy. Comprehensive ablation studies further demonstrate its robustness across challenging conditions. |
Accep...Accepted to CVPR 2025 |
| Open-Vocabulary Semantic Segmentation with Uncertainty Alignment for Robotic Scene Understanding in Indoor Building Environments | 2025-03-29 | ShowThe global rise in the number of people with physical disabilities, in part due to improvements in post-trauma survivorship and longevity, has amplified the demand for advanced assistive technologies to improve mobility and independence. Autonomous assistive robots, such as smart wheelchairs, require robust capabilities in spatial segmentation and semantic recognition to navigate complex built environments effectively. Place segmentation involves delineating spatial regions like rooms or functional areas, while semantic recognition assigns semantic labels to these regions, enabling accurate localization to user-specific needs. Existing approaches often utilize deep learning; however, these close-vocabulary detection systems struggle to interpret intuitive and casual human instructions. Additionally, most existing methods ignore the uncertainty of the scene recognition problem, leading to low success rates, particularly in ambiguous and complex environments. To address these challenges, we propose an open-vocabulary scene semantic segmentation and detection pipeline leveraging Vision Language Models (VLMs) and Large Language Models (LLMs). Our approach follows a 'Segment Detect Select' framework for open-vocabulary scene classification, enabling adaptive and intuitive navigation for assistive robots in built environments. |
32 pages, 7 figures |
| Semantic Library Adaptation: LoRA Retrieval and Fusion for Open-Vocabulary Semantic Segmentation | 2025-03-27 | ShowOpen-vocabulary semantic segmentation models associate vision and text to label pixels from an undefined set of classes using textual queries, providing versatile performance on novel datasets. However, large shifts between training and test domains degrade their performance, requiring fine-tuning for effective real-world applications. We introduce Semantic Library Adaptation (SemLA), a novel framework for training-free, test-time domain adaptation. SemLA leverages a library of LoRA-based adapters indexed with CLIP embeddings, dynamically merging the most relevant adapters based on proximity to the target domain in the embedding space. This approach constructs an ad-hoc model tailored to each specific input without additional training. Our method scales efficiently, enhances explainability by tracking adapter contributions, and inherently protects data privacy, making it ideal for sensitive applications. Comprehensive experiments on a 20-domain benchmark built over 10 standard datasets demonstrate SemLA's superior adaptability and performance across diverse settings, establishing a new standard in domain adaptation for open-vocabulary semantic segmentation. |
CVPR ...CVPR 2025. Project page: https://thegoodailab.org/semla Code: https://github.com/rezaqorbani/SemLA |
| Rethinking the Global Knowledge of CLIP in Training-Free Open-Vocabulary Semantic Segmentation | 2025-03-27 | ShowRecent works modify CLIP to perform open-vocabulary semantic segmentation in a training-free manner (TF-OVSS). In vanilla CLIP, patch-wise image representations mainly encode homogeneous image-level properties, which hinders the application of CLIP to the dense prediction task. Previous TF-OVSS works sacrifice globality to enhance the locality of CLIP features, by making each patch mainly attend to itself or its neighboring patches within a narrow local window. With their modifications,the ability of CLIP to aggregate global context information is largely weakened. Differently, in this paper, we rethink the global knowledge encoded by CLIP and propose GCLIP to answer how to extract and utilize beneficial global knowledge of CLIP for TF-OVSS. As the representation of each patch is finally determined by the attention weights and the Value embeddings, we propose to reshape the last-block attention and Value embeddings to aggregate useful global context into final features. Firstly, we aim to equip the last-block attention with image-level properties while not introducing homogeneous attention patterns across patches. To realize the goal, we fuse the attention from the global-token emerging blocks with the Query-Query attention. Secondly, we aim to make Value embeddings of the last-block attention module more semantically correlated. To realize this, we design a novel channel suppression strategy.Extensive experiments on five standard benchmarks demonstrate that our method consistently outperforms previous state-of-the-arts. |
Under review |
| Test-Time Optimization for Domain Adaptive Open Vocabulary Segmentation | 2025-03-08 | ShowWe present Seg-TTO, a novel framework for zero-shot, open-vocabulary semantic segmentation (OVSS), designed to excel in specialized domain tasks. While current open-vocabulary approaches show impressive performance on standard segmentation benchmarks under zero-shot settings, they fall short of supervised counterparts on highly domain-specific datasets. We focus on segmentation-specific test-time optimization to address this gap. Segmentation requires an understanding of multiple concepts within a single image while retaining the locality and spatial structure of representations. We propose a novel self-supervised objective adhering to these requirements and use it to align the model parameters with input images at test time. In the textual modality, we learn multiple embeddings for each category to capture diverse concepts within an image, while in the visual modality, we calculate pixel-level losses followed by embedding aggregation operations specific to preserving spatial structure. Our resulting framework termed Seg-TTO is a plug-and-play module. We integrate Seg-TTO with three state-of-the-art OVSS approaches and evaluate across 22 challenging OVSS tasks covering a range of specialized domains. Our Seg-TTO demonstrates clear performance improvements (up to 27% mIoU increase on some datasets) establishing new state-of-the-art. Our code and models will be released publicly. |
|
| From Open-Vocabulary to Vocabulary-Free Semantic Segmentation | 2025-02-17 | ShowOpen-vocabulary semantic segmentation enables models to identify novel object categories beyond their training data. While this flexibility represents a significant advancement, current approaches still rely on manually specified class names as input, creating an inherent bottleneck in real-world applications. This work proposes a Vocabulary-Free Semantic Segmentation pipeline, eliminating the need for predefined class vocabularies. Specifically, we address the chicken-and-egg problem where users need knowledge of all potential objects within a scene to identify them, yet the purpose of segmentation is often to discover these objects. The proposed approach leverages Vision-Language Models to automatically recognize objects and generate appropriate class names, aiming to solve the challenge of class specification and naming quality. Through extensive experiments on several public datasets, we highlight the crucial role of the text encoder in model performance, particularly when the image text classes are paired with generated descriptions. Despite the challenges introduced by the sensitivity of the segmentation text encoder to false negatives within the class tagging process, which adds complexity to the task, we demonstrate that our fully automated pipeline significantly enhances vocabulary-free segmentation accuracy across diverse real-world scenarios. |
Submi...Submitted to: Pattern Recognition Letters, Klara Reichard and Giulia Rizzoli equally contributed to this work |
| InvSeg: Test-Time Prompt Inversion for Semantic Segmentation | 2025-01-03 | ShowVisual-textual correlations in the attention maps derived from text-to-image diffusion models are proven beneficial to dense visual prediction tasks, e.g., semantic segmentation. However, a significant challenge arises due to the input distributional discrepancy between the context-rich sentences used for image generation and the isolated class names typically used in semantic segmentation. This discrepancy hinders diffusion models from capturing accurate visual-textual correlations. To solve this, we propose InvSeg, a test-time prompt inversion method that tackles open-vocabulary semantic segmentation by inverting image-specific visual context into text prompt embedding space, leveraging structure information derived from the diffusion model's reconstruction process to enrich text prompts so as to associate each class with a structure-consistent mask. Specifically, we introduce Contrastive Soft Clustering (CSC) to align derived masks with the image's structure information, softly selecting anchors for each class and calculating weighted distances to push inner-class pixels closer while separating inter-class pixels, thereby ensuring mask distinction and internal consistency. By incorporating sample-specific context, InvSeg learns context-rich text prompts in embedding space and achieves accurate semantic alignment across modalities. Experiments show that InvSeg achieves state-of-the-art performance on the PASCAL VOC, PASCAL Context and COCO Object datasets. |
AAAI 2025 |
| FGAseg: Fine-Grained Pixel-Text Alignment for Open-Vocabulary Semantic Segmentation | 2025-01-03 | ShowOpen-vocabulary segmentation aims to identify and segment specific regions and objects based on text-based descriptions. A common solution is to leverage powerful vision-language models (VLMs), such as CLIP, to bridge the gap between vision and text information. However, VLMs are typically pretrained for image-level vision-text alignment, focusing on global semantic features. In contrast, segmentation tasks require fine-grained pixel-level alignment and detailed category boundary information, which VLMs alone cannot provide. As a result, information extracted directly from VLMs can't meet the requirements of segmentation tasks. To address this limitation, we propose FGAseg, a model designed for fine-grained pixel-text alignment and category boundary supplementation. The core of FGAseg is a Pixel-Level Alignment module that employs a cross-modal attention mechanism and a text-pixel alignment loss to refine the coarse-grained alignment from CLIP, achieving finer-grained pixel-text semantic alignment. Additionally, to enrich category boundary information, we introduce the alignment matrices as optimizable pseudo-masks during forward propagation and propose Category Information Supplementation module. These pseudo-masks, derived from cosine and convolutional similarity, provide essential global and local boundary information between different categories. By combining these two strategies, FGAseg effectively enhances pixel-level alignment and category boundary information, addressing key challenges in open-vocabulary segmentation. Extensive experiments demonstrate that FGAseg outperforms existing methods on open-vocabulary semantic segmentation benchmarks. |
|
| OVGaussian: Generalizable 3D Gaussian Segmentation with Open Vocabularies | 2024-12-31 | ShowOpen-vocabulary scene understanding using 3D Gaussian (3DGS) representations has garnered considerable attention. However, existing methods mostly lift knowledge from large 2D vision models into 3DGS on a scene-by-scene basis, restricting the capabilities of open-vocabulary querying within their training scenes so that lacking the generalizability to novel scenes. In this work, we propose \textbf{OVGaussian}, a generalizable \textbf{O}pen-\textbf{V}ocabulary 3D semantic segmentation framework based on the 3D \textbf{Gaussian} representation. We first construct a large-scale 3D scene dataset based on 3DGS, dubbed \textbf{SegGaussian}, which provides detailed semantic and instance annotations for both Gaussian points and multi-view images. To promote semantic generalization across scenes, we introduce Generalizable Semantic Rasterization (GSR), which leverages a 3D neural network to learn and predict the semantic property for each 3D Gaussian point, where the semantic property can be rendered as multi-view consistent 2D semantic maps. In the next, we propose a Cross-modal Consistency Learning (CCL) framework that utilizes open-vocabulary annotations of 2D images and 3D Gaussians within SegGaussian to train the 3D neural network capable of open-vocabulary semantic segmentation across Gaussian-based 3D scenes. Experimental results demonstrate that OVGaussian significantly outperforms baseline methods, exhibiting robust cross-scene, cross-domain, and novel-view generalization capabilities. Code and the SegGaussian dataset will be released. (https://github.com/runnanchen/OVGaussian). |
|
| DINOv2 Meets Text: A Unified Framework for Image- and Pixel-Level Vision-Language Alignment | 2024-12-20 | ShowSelf-supervised visual foundation models produce powerful embeddings that achieve remarkable performance on a wide range of downstream tasks. However, unlike vision-language models such as CLIP, self-supervised visual features are not readily aligned with language, hindering their adoption in open-vocabulary tasks. Our method, named dino.txt, unlocks this new ability for DINOv2, a widely used self-supervised visual encoder. We build upon the LiT training strategy, which trains a text encoder to align with a frozen vision model but leads to unsatisfactory results on dense tasks. We propose several key ingredients to improve performance on both global and dense tasks, such as concatenating the [CLS] token with the patch average to train the alignment and curating data using both text and image modalities. With these, we successfully train a CLIP-like model with only a fraction of the computational cost compared to CLIP while achieving state-of-the-art results in zero-shot classification and open-vocabulary semantic segmentation. |
|
| Incorporating Feature Pyramid Tokenization and Open Vocabulary Semantic Segmentation | 2024-12-18 | ShowThe visual understanding are often approached from 3 granular levels: image, patch and pixel. Visual Tokenization, trained by self-supervised reconstructive learning, compresses visual data by codebook in patch-level with marginal information loss, but the visual tokens does not have semantic meaning. Open Vocabulary semantic segmentation benefits from the evolving Vision-Language models (VLMs) with strong image zero-shot capability, but transferring image-level to pixel-level understanding remains an imminent challenge. In this paper, we treat segmentation as tokenizing pixels and study a united perceptual and semantic token compression for all granular understanding and consequently facilitate open vocabulary semantic segmentation. Referring to the cognitive process of pretrained VLM where the low-level features are progressively composed to high-level semantics, we propose Feature Pyramid Tokenization (PAT) to cluster and represent multi-resolution feature by learnable codebooks and then decode them by joint learning pixel reconstruction and semantic segmentation. We design loosely coupled pixel and semantic learning branches. The pixel branch simulates bottom-up composition and top-down visualization of codebook tokens, while the semantic branch collectively fuse hierarchical codebooks as auxiliary segmentation guidance. Our experiments show that PAT enhances the semantic intuition of VLM feature pyramid, improves performance over the baseline segmentation model and achieves competitive performance on open vocabulary semantic segmentation benchmark. Our model is parameter-efficient for VLM integration and flexible for the independent tokenization. We hope to give inspiration not only on improving segmentation but also on semantic visual token utilization. |
6 pages, 6 figures |
| VLMs meet UDA: Boosting Transferability of Open Vocabulary Segmentation with Unsupervised Domain Adaptation | 2024-12-12 | ShowSegmentation models are typically constrained by the categories defined during training. To address this, researchers have explored two independent approaches: adapting Vision-Language Models (VLMs) and leveraging synthetic data. However, VLMs often struggle with granularity, failing to disentangle fine-grained concepts, while synthetic data-based methods remain limited by the scope of available datasets. This paper proposes enhancing segmentation accuracy across diverse domains by integrating Vision-Language reasoning with key strategies for Unsupervised Domain Adaptation (UDA). First, we improve the fine-grained segmentation capabilities of VLMs through multi-scale contextual data, robust text embeddings with prompt augmentation, and layer-wise fine-tuning in our proposed Foundational-Retaining Open Vocabulary Semantic Segmentation (FROVSS) framework. Next, we incorporate these enhancements into a UDA framework by employing distillation to stabilize training and cross-domain mixed sampling to boost adaptability without compromising generalization. The resulting UDA-FROVSS framework is the first UDA approach to effectively adapt across domains without requiring shared categories. |
|
| Collaborative Vision-Text Representation Optimizing for Open-Vocabulary Segmentation | 2024-12-04 | ShowPre-trained vision-language models, e.g. CLIP, have been increasingly used to address the challenging Open-Vocabulary Segmentation (OVS) task, benefiting from their well-aligned vision-text embedding space. Typical solutions involve either freezing CLIP during training to unilaterally maintain its zero-shot capability, or fine-tuning CLIP vision encoder to achieve perceptual sensitivity to local regions. However, few of them incorporate vision-text collaborative optimization. Based on this, we propose the Content-Dependent Transfer to adaptively enhance each text embedding by interacting with the input image, which presents a parameter-efficient way to optimize the text representation. Besides, we additionally introduce a Representation Compensation strategy, reviewing the original CLIP-V representation as compensation to maintain the zero-shot capability of CLIP. In this way, the vision and text representation of CLIP are optimized collaboratively, enhancing the alignment of the vision-text feature space. To the best of our knowledge, we are the first to establish the collaborative vision-text optimizing mechanism within the OVS field. Extensive experiments demonstrate our method achieves superior performance on popular OVS benchmarks. In open-vocabulary semantic segmentation, our method outperforms the previous state-of-the-art approaches by +0.5, +2.3, +3.4, +0.4 and +1.1 mIoU, respectively on A-847, A-150, PC-459, PC-59 and PAS-20. Furthermore, in a panoptic setting on ADE20K, we achieve the performance of 27.1 PQ, 73.5 SQ, and 32.9 RQ. Code will be available at https://github.com/jiaosiyu1999/MAFT-Plus.git . |
ECCV 2024 oral |
| Multi-robot autonomous 3D reconstruction using Gaussian splatting with Semantic guidance | 2024-12-03 | ShowImplicit neural representations and 3D Gaussian splatting (3DGS) have shown great potential for scene reconstruction. Recent studies have expanded their applications in autonomous reconstruction through task assignment methods. However, these methods are mainly limited to single robot, and rapid reconstruction of large-scale scenes remains challenging. Additionally, task-driven planning based on surface uncertainty is prone to being trapped in local optima. To this end, we propose the first 3DGS-based centralized multi-robot autonomous 3D reconstruction framework. To further reduce time cost of task generation and improve reconstruction quality, we integrate online open-vocabulary semantic segmentation with surface uncertainty of 3DGS, focusing view sampling on regions with high instance uncertainty. Finally, we develop a multi-robot collaboration strategy with mode and task assignments improving reconstruction quality while ensuring planning efficiency. Our method demonstrates the highest reconstruction quality among all planning methods and superior planning efficiency compared to existing multi-robot methods. We deploy our method on multiple robots, and results show that it can effectively plan view paths and reconstruct scenes with high quality. |
|
| Parameter-efficient Fine-tuning in Hyperspherical Space for Open-vocabulary Semantic Segmentation | 2024-11-29 | ShowOpen-vocabulary semantic segmentation seeks to label each pixel in an image with arbitrary text descriptions. Vision-language foundation models, especially CLIP, have recently emerged as powerful tools for acquiring open-vocabulary capabilities. However, fine-tuning CLIP to equip it with pixel-level prediction ability often suffers three issues: 1) high computational cost, 2) misalignment between the two inherent modalities of CLIP, and 3) degraded generalization ability on unseen categories. To address these issues, we propose H-CLIP a symmetrical parameter-efficient fine-tuning (PEFT) strategy conducted in hyperspherical space for both of the two CLIP modalities. Specifically, the PEFT strategy is achieved by a series of efficient block-diagonal learnable transformation matrices and a dual cross-relation communication module among all learnable matrices. Since the PEFT strategy is conducted symmetrically to the two CLIP modalities, the misalignment between them is mitigated. Furthermore, we apply an additional constraint to PEFT on the CLIP text encoder according to the hyperspherical energy principle, i.e., minimizing hyperspherical energy during fine-tuning preserves the intrinsic structure of the original parameter space, to prevent the destruction of the generalization ability offered by the CLIP text encoder. Extensive evaluations across various benchmarks show that H-CLIP achieves new SOTA open-vocabulary semantic segmentation results while only requiring updating approximately 4% of the total parameters of CLIP. |
|
| CLIPer: Hierarchically Improving Spatial Representation of CLIP for Open-Vocabulary Semantic Segmentation | 2024-11-21 | ShowContrastive Language-Image Pre-training (CLIP) exhibits strong zero-shot classification ability on various image-level tasks, leading to the research to adapt CLIP for pixel-level open-vocabulary semantic segmentation without additional training. The key is to improve spatial representation of image-level CLIP, such as replacing self-attention map at last layer with self-self attention map or vision foundation model based attention map. In this paper, we present a novel hierarchical framework, named CLIPer, that hierarchically improves spatial representation of CLIP. The proposed CLIPer includes an early-layer fusion module and a fine-grained compensation module. We observe that, the embeddings and attention maps at early layers can preserve spatial structural information. Inspired by this, we design the early-layer fusion module to generate segmentation map with better spatial coherence. Afterwards, we employ a fine-grained compensation module to compensate the local details using the self-attention maps of diffusion model. We conduct the experiments on seven segmentation datasets. Our proposed CLIPer achieves the state-of-the-art performance on these datasets. For instance, using ViT-L, CLIPer has the mIoU of 69.8% and 43.3% on VOC and COCO Object, outperforming ProxyCLIP by 9.2% and 4.1% respectively. |
Homep...Homepange and code: https://linsun449.github.io/cliper |
| XMask3D: Cross-modal Mask Reasoning for Open Vocabulary 3D Semantic Segmentation | 2024-11-20 | ShowExisting methodologies in open vocabulary 3D semantic segmentation primarily concentrate on establishing a unified feature space encompassing 3D, 2D, and textual modalities. Nevertheless, traditional techniques such as global feature alignment or vision-language model distillation tend to impose only approximate correspondence, struggling notably with delineating fine-grained segmentation boundaries. To address this gap, we propose a more meticulous mask-level alignment between 3D features and the 2D-text embedding space through a cross-modal mask reasoning framework, XMask3D. In our approach, we developed a mask generator based on the denoising UNet from a pre-trained diffusion model, leveraging its capability for precise textual control over dense pixel representations and enhancing the open-world adaptability of the generated masks. We further integrate 3D global features as implicit conditions into the pre-trained 2D denoising UNet, enabling the generation of segmentation masks with additional 3D geometry awareness. Subsequently, the generated 2D masks are employed to align mask-level 3D representations with the vision-language feature space, thereby augmenting the open vocabulary capability of 3D geometry embeddings. Finally, we fuse complementary 2D and 3D mask features, resulting in competitive performance across multiple benchmarks for 3D open vocabulary semantic segmentation. Code is available at https://github.com/wangzy22/XMask3D. |
Accep...Accepted to NeurIPS 2024 |
| SegEarth-OV: Towards Training-Free Open-Vocabulary Segmentation for Remote Sensing Images | 2024-11-04 | ShowRemote sensing image plays an irreplaceable role in fields such as agriculture, water resources, military, and disaster relief. Pixel-level interpretation is a critical aspect of remote sensing image applications; however, a prevalent limitation remains the need for extensive manual annotation. For this, we try to introduce open-vocabulary semantic segmentation (OVSS) into the remote sensing context. However, due to the sensitivity of remote sensing images to low-resolution features, distorted target shapes and ill-fitting boundaries are exhibited in the prediction mask. To tackle this issue, we propose a simple and general upsampler, SimFeatUp, to restore lost spatial information in deep features in a training-free style. Further, based on the observation of the abnormal response of local patch tokens to [CLS] token in CLIP, we propose to execute a straightforward subtraction operation to alleviate the global bias in patch tokens. Extensive experiments are conducted on 17 remote sensing datasets spanning semantic segmentation, building extraction, road detection, and flood detection tasks. Our method achieves an average of 5.8%, 8.2%, 4.0%, and 15.3% improvement over state-of-the-art methods on 4 tasks. All codes are released. \url{https://earth-insights.github.io/SegEarth-OV} |
|
| EAGLE: Efficient Adaptive Geometry-based Learning in Cross-view Understanding | 2024-10-11 | ShowUnsupervised Domain Adaptation has been an efficient approach to transferring the semantic segmentation model across data distributions. Meanwhile, the recent Open-vocabulary Semantic Scene understanding based on large-scale vision language models is effective in open-set settings because it can learn diverse concepts and categories. However, these prior methods fail to generalize across different camera views due to the lack of cross-view geometric modeling. At present, there are limited studies analyzing cross-view learning. To address this problem, we introduce a novel Unsupervised Cross-view Adaptation Learning approach to modeling the geometric structural change across views in Semantic Scene Understanding. First, we introduce a novel Cross-view Geometric Constraint on Unpaired Data to model structural changes in images and segmentation masks across cameras. Second, we present a new Geodesic Flow-based Correlation Metric to efficiently measure the geometric structural changes across camera views. Third, we introduce a novel view-condition prompting mechanism to enhance the view-information modeling of the open-vocabulary segmentation network in cross-view adaptation learning. The experiments on different cross-view adaptation benchmarks have shown the effectiveness of our approach in cross-view modeling, demonstrating that we achieve State-of-the-Art (SOTA) performance compared to prior unsupervised domain adaptation and open-vocabulary semantic segmentation methods. |
Accep...Accepted to NeurIPS'24 |
| Open-RGBT: Open-vocabulary RGB-T Zero-shot Semantic Segmentation in Open-world Environments | 2024-10-09 | ShowSemantic segmentation is a critical technique for effective scene understanding. Traditional RGB-T semantic segmentation models often struggle to generalize across diverse scenarios due to their reliance on pretrained models and predefined categories. Recent advancements in Visual Language Models (VLMs) have facilitated a shift from closed-set to open-vocabulary semantic segmentation methods. However, these models face challenges in dealing with intricate scenes, primarily due to the heterogeneity between RGB and thermal modalities. To address this gap, we present Open-RGBT, a novel open-vocabulary RGB-T semantic segmentation model. Specifically, we obtain instance-level detection proposals by incorporating visual prompts to enhance category understanding. Additionally, we employ the CLIP model to assess image-text similarity, which helps correct semantic consistency and mitigates ambiguities in category identification. Empirical evaluations demonstrate that Open-RGBT achieves superior performance in diverse and challenging real-world scenarios, even in the wild, significantly advancing the field of RGB-T semantic segmentation. |
|
| Towards Open-Vocabulary Semantic Segmentation Without Semantic Labels | 2024-09-30 | ShowLarge-scale vision-language models like CLIP have demonstrated impressive open-vocabulary capabilities for image-level tasks, excelling in recognizing what objects are present. However, they struggle with pixel-level recognition tasks like semantic segmentation, which additionally require understanding where the objects are located. In this work, we propose a novel method, PixelCLIP, to adapt the CLIP image encoder for pixel-level understanding by guiding the model on where, which is achieved using unlabeled images and masks generated from vision foundation models such as SAM and DINO. To address the challenges of leveraging masks without semantic labels, we devise an online clustering algorithm using learnable class names to acquire general semantic concepts. PixelCLIP shows significant performance improvements over CLIP and competitive results compared to caption-supervised methods in open-vocabulary semantic segmentation. Project page is available at https://cvlab-kaist.github.io/PixelCLIP |
To ap...To appear at NeurIPS 2024. Project page is available at https://cvlab-kaist.github.io/PixelCLIP |
| CUS3D :CLIP-based Unsupervised 3D Segmentation via Object-level Denoise | 2024-09-21 | ShowTo ease the difficulty of acquiring annotation labels in 3D data, a common method is using unsupervised and open-vocabulary semantic segmentation, which leverage 2D CLIP semantic knowledge. In this paper, unlike previous research that ignores the |
6 pages,3 figures |
| Pay Attention to Your Neighbours: Training-Free Open-Vocabulary Semantic Segmentation | 2024-09-16 | ShowDespite the significant progress in deep learning for dense visual recognition problems, such as semantic segmentation, traditional methods are constrained by fixed class sets. Meanwhile, vision-language foundation models, such as CLIP, have showcased remarkable effectiveness in numerous zero-shot image-level tasks, owing to their robust generalizability. Recently, a body of work has investigated utilizing these models in open-vocabulary semantic segmentation (OVSS). However, existing approaches often rely on impractical supervised pre-training or access to additional pre-trained networks. In this work, we propose a strong baseline for training-free OVSS, termed Neighbour-Aware CLIP (NACLIP), representing a straightforward adaptation of CLIP tailored for this scenario. Our method enforces localization of patches in the self-attention of CLIP's vision transformer which, despite being crucial for dense prediction tasks, has been overlooked in the OVSS literature. By incorporating design choices favouring segmentation, our approach significantly improves performance without requiring additional data, auxiliary pre-trained networks, or extensive hyperparameter tuning, making it highly practical for real-world applications. Experiments are performed on 8 popular semantic segmentation benchmarks, yielding state-of-the-art performance on most scenarios. Our code is publicly available at https://github.com/sinahmr/NACLIP. |
Accep...Accepted to WACV 2025 |
| TagCLIP: Improving Discrimination Ability of Open-Vocabulary Semantic Segmentation | 2024-09-03 | ShowContrastive Language-Image Pre-training (CLIP) has recently shown great promise in pixel-level zero-shot learning tasks. However, existing approaches utilizing CLIP's text and patch embeddings to generate semantic masks often misidentify input pixels from unseen classes, leading to confusion between novel classes and semantically similar ones. In this work, we propose a novel approach, TagCLIP (Trusty-aware guided CLIP), to address this issue. We disentangle the ill-posed optimization problem into two parallel processes: semantic matching performed individually and reliability judgment for improving discrimination ability. Building on the idea of special tokens in language modeling representing sentence-level embeddings, we introduce a trusty token that enables distinguishing novel classes from known ones in prediction. To evaluate our approach, we conduct experiments on two benchmark datasets, PASCAL VOC 2012, COCO-Stuff 164K and PASCAL Context. Our results show that TagCLIP improves the Intersection over Union (IoU) of unseen classes by 7.4%, 1.7% and 2.1%, respectively, with negligible overheads. The code is available at https://github.com/dvlab-research/TagCLIP. |
TPAMI2024 |
| OVOSE: Open-Vocabulary Semantic Segmentation in Event-Based Cameras | 2024-08-18 | ShowEvent cameras, known for low-latency operation and superior performance in challenging lighting conditions, are suitable for sensitive computer vision tasks such as semantic segmentation in autonomous driving. However, challenges arise due to limited event-based data and the absence of large-scale segmentation benchmarks. Current works are confined to closed-set semantic segmentation, limiting their adaptability to other applications. In this paper, we introduce OVOSE, the first Open-Vocabulary Semantic Segmentation algorithm for Event cameras. OVOSE leverages synthetic event data and knowledge distillation from a pre-trained image-based foundation model to an event-based counterpart, effectively preserving spatial context and transferring open-vocabulary semantic segmentation capabilities. We evaluate the performance of OVOSE on two driving semantic segmentation datasets DDD17, and DSEC-Semantic, comparing it with existing conventional image open-vocabulary models adapted for event-based data. Similarly, we compare OVOSE with state-of-the-art methods designed for closed-set settings in unsupervised domain adaptation for event-based semantic segmentation. OVOSE demonstrates superior performance, showcasing its potential for real-world applications. The code is available at https://github.com/ram95d/OVOSE. |
conference |
| In Defense of Lazy Visual Grounding for Open-Vocabulary Semantic Segmentation | 2024-08-09 | ShowWe present lazy visual grounding, a two-stage approach of unsupervised object mask discovery followed by object grounding, for open-vocabulary semantic segmentation. Plenty of the previous art casts this task as pixel-to-text classification without object-level comprehension, leveraging the image-to-text classification capability of pretrained vision-and-language models. We argue that visual objects are distinguishable without the prior text information as segmentation is essentially a vision task. Lazy visual grounding first discovers object masks covering an image with iterative Normalized cuts and then later assigns text on the discovered objects in a late interaction manner. Our model requires no additional training yet shows great performance on five public datasets: Pascal VOC, Pascal Context, COCO-object, COCO-stuff, and ADE 20K. Especially, the visually appealing segmentation results demonstrate the model capability to localize objects precisely. Paper homepage: https://cvlab.postech.ac.kr/research/lazygrounding |
Accep...Accepted to ECCV 2024 |
| ProxyCLIP: Proxy Attention Improves CLIP for Open-Vocabulary Segmentation | 2024-08-09 | ShowOpen-vocabulary semantic segmentation requires models to effectively integrate visual representations with open-vocabulary semantic labels. While Contrastive Language-Image Pre-training (CLIP) models shine in recognizing visual concepts from text, they often struggle with segment coherence due to their limited localization ability. In contrast, Vision Foundation Models (VFMs) excel at acquiring spatially consistent local visual representations, yet they fall short in semantic understanding. This paper introduces ProxyCLIP, an innovative framework designed to harmonize the strengths of both CLIP and VFMs, facilitating enhanced open-vocabulary semantic segmentation. ProxyCLIP leverages the spatial feature correspondence from VFMs as a form of proxy attention to augment CLIP, thereby inheriting the VFMs' robust local consistency and maintaining CLIP's exceptional zero-shot transfer capacity. We propose an adaptive normalization and masking strategy to get the proxy attention from VFMs, allowing for adaptation across different VFMs. Remarkably, as a training-free approach, ProxyCLIP significantly improves the average mean Intersection over Union (mIoU) across eight benchmarks from 40.3 to 44.4, showcasing its exceptional efficacy in bridging the gap between spatial precision and semantic richness for the open-vocabulary segmentation task. |
Accep...Accepted to ECCV 2024. Code available at https://github.com/mc-lan/ProxyCLIP |
| SegPoint: Segment Any Point Cloud via Large Language Model | 2024-07-18 | ShowDespite significant progress in 3D point cloud segmentation, existing methods primarily address specific tasks and depend on explicit instructions to identify targets, lacking the capability to infer and understand implicit user intentions in a unified framework. In this work, we propose a model, called SegPoint, that leverages the reasoning capabilities of a multi-modal Large Language Model (LLM) to produce point-wise segmentation masks across a diverse range of tasks: 1) 3D instruction segmentation, 2) 3D referring segmentation, 3) 3D semantic segmentation, and 4) 3D open-vocabulary semantic segmentation. To advance 3D instruction research, we introduce a new benchmark, Instruct3D, designed to evaluate segmentation performance from complex and implicit instructional texts, featuring 2,565 point cloud-instruction pairs. Our experimental results demonstrate that SegPoint achieves competitive performance on established benchmarks such as ScanRefer for referring segmentation and ScanNet for semantic segmentation, while delivering outstanding outcomes on the Instruct3D dataset. To our knowledge, SegPoint is the first model to address these varied segmentation tasks within a single framework, achieving satisfactory performance. |
ECCV ...ECCV 2024, Project Page: https://heshuting555.github.io/SegPoint |
| ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inference | 2024-07-17 | ShowDespite the success of large-scale pretrained Vision-Language Models (VLMs) especially CLIP in various open-vocabulary tasks, their application to semantic segmentation remains challenging, producing noisy segmentation maps with mis-segmented regions. In this paper, we carefully re-investigate the architecture of CLIP, and identify residual connections as the primary source of noise that degrades segmentation quality. With a comparative analysis of statistical properties in the residual connection and the attention output across different pretrained models, we discover that CLIP's image-text contrastive training paradigm emphasizes global features at the expense of local discriminability, leading to noisy segmentation results. In response, we propose ClearCLIP, a novel approach that decomposes CLIP's representations to enhance open-vocabulary semantic segmentation. We introduce three simple modifications to the final layer: removing the residual connection, implementing the self-self attention, and discarding the feed-forward network. ClearCLIP consistently generates clearer and more accurate segmentation maps and outperforms existing approaches across multiple benchmarks, affirming the significance of our discoveries. |
Accep...Accepted to ECCV 2024. code available at https://github.com/mc- lan/ClearCLIP |
| Open Panoramic Segmentation | 2024-07-11 | ShowPanoramic images, capturing a 360° field of view (FoV), encompass omnidirectional spatial information crucial for scene understanding. However, it is not only costly to obtain training-sufficient dense-annotated panoramas but also application-restricted when training models in a close-vocabulary setting. To tackle this problem, in this work, we define a new task termed Open Panoramic Segmentation (OPS), where models are trained with FoV-restricted pinhole images in the source domain in an open-vocabulary setting while evaluated with FoV-open panoramic images in the target domain, enabling the zero-shot open panoramic semantic segmentation ability of models. Moreover, we propose a model named OOOPS with a Deformable Adapter Network (DAN), which significantly improves zero-shot panoramic semantic segmentation performance. To further enhance the distortion-aware modeling ability from the pinhole source domain, we propose a novel data augmentation method called Random Equirectangular Projection (RERP) which is specifically designed to address object deformations in advance. Surpassing other state-of-the-art open-vocabulary semantic segmentation approaches, a remarkable performance boost on three panoramic datasets, WildPASS, Stanford2D3D, and Matterport3D, proves the effectiveness of our proposed OOOPS model with RERP on the OPS task, especially +2.2% on outdoor WildPASS and +2.4% mIoU on indoor Stanford2D3D. The source code is publicly available at https://junweizheng93.github.io/publications/OPS/OPS.html. |
Accep...Accepted by ECCV 2024. Project page: https://junweizheng93.github.io/publications/OPS/OPS.html |
| Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation | 2024-07-11 | ShowCLIP, as a vision-language model, has significantly advanced Open-Vocabulary Semantic Segmentation (OVSS) with its zero-shot capabilities. Despite its success, its application to OVSS faces challenges due to its initial image-level alignment training, which affects its performance in tasks requiring detailed local context. Our study delves into the impact of CLIP's [CLS] token on patch feature correlations, revealing a dominance of "global" patches that hinders local feature discrimination. To overcome this, we propose CLIPtrase, a novel training-free semantic segmentation strategy that enhances local feature awareness through recalibrated self-correlation among patches. This approach demonstrates notable improvements in segmentation accuracy and the ability to maintain semantic coherence across objects.Experiments show that we are 22.3% ahead of CLIP on average on 9 segmentation benchmarks, outperforming existing state-of-the-art training-free methods.The code are made publicly available at: https://github.com/leaves162/CLIPtrase. |
ECCV24 accepted |
| Title | Date | Abstract | Comment |
|---|---|---|---|
| CoCo-SAM3: Harnessing Concept Conflict in Open-Vocabulary Semantic Segmentation | 2026-04-21 | ShowSAM3 advances open-vocabulary semantic segmentation by introducing a prompt-driven mask generation paradigm. However, in multi-class open-vocabulary scenarios, masks generated independently from different category prompts lack a unified and inter-class comparable evidence scale, often resulting in overlapping coverage and unstable competition. Moreover, synonymous expressions of the same concept tend to activate inconsistent semantic and spatial evidence, leading to intra-class drift that exacerbates inter-class conflicts and compromises overall inference stability. To address these issues, we propose CoCo-SAM3 (Concept-Conflict SAM3), which explicitly decouples inference into intra-class enhancement and inter-class competition. Our method first aligns and aggregates evidence from synonymous prompts to strengthen concept consistency. It then performs inter-class competition on a unified comparable scale, enabling direct pixel-wise comparisons among all candidate classes. This mechanism stabilizes multi-class inference and effectively mitigates inter-class conflicts. Without requiring any additional training, CoCo-SAM3 achieves consistent improvements across eight open-vocabulary semantic segmentation benchmarks. |
|
| Is SAM3 ready for pathology segmentation? | 2026-04-20 | ShowIs Segment Anything Model 3 (SAM3) capable in segmenting Any Pathology Images? Digital pathology segmentation spans tissue-level and nuclei-level scales, where traditional methods often suffer from high annotation costs and poor generalization. SAM3 introduces Promptable Concept Segmentation, offering a potential automated interface via text prompts. With this work, we propose a systematic evaluation protocol to explore the capability space of SAM3 in a structured manner. Specifically, we evaluate SAM3 under different supervision settings including zero-shot, few-shot, and supervised with varying prompting strategies. Our extensive evaluation on pathological datasets including NuInsSeg, PanNuke and GlaS, reveals that: 1.text-only prompts poorly activate nuclear concepts. 2.performance is highly sensitive to visual prompt types and budgets. 3.few-shot learning offers gains, but SAM3 lacks robustness against visual prompt noise. and 4.a significant gap persists between prompt-based usage and task-trained adapter-based reference. Our study delineates SAM3's boundaries in pathology image segmentation and provides practical guidance on the necessity of pathology domain adaptation. |
|
| SAM3-I: Segment Anything with Instructions | 2026-04-16 | ShowSegment Anything Model 3 (SAM3) advances open-vocabulary segmentation through promptable concept segmentation, enabling users to segment all instances associated with a given concept using short noun-phrase (NP) prompts. While effective for concept-level grounding, real-world interactions often involve far richer natural-language instructions that combine attributes, relations, actions, states, or implicit reasoning. Currently, SAM3 relies on external multi-modal agents to convert complex instructions into NPs and conducts iterative mask filtering, leading to coarse representations and limited instance specificity. In this work, we present SAM3-I, an instruction-following extension of the SAM family that unifies concept-level grounding and instruction-level reasoning within a single segmentation framework. Built upon SAM3, SAM3-I introduces an instruction-aware cascaded adaptation mechanism with dedicated alignment losses that progressively aligns expressive instruction semantics with SAM3's vision-language representations, enabling direct interpretation of natural-language instructions while preserving its strong concept recall ability. To enable instruction-following learning, we introduce HMPL-Instruct, a large-scale instruction-centric dataset that systematically covers hierarchical instruction semantics and diverse target granularities. Experiments demonstrate that SAM3-I achieves appealing performance across referring and reasoning-based segmentation, showing that SAM3 can be effectively extended to follow complex natural-language instructions without sacrificing its original concept-driven strengths. Code and dataset are available at https://github.com/debby-0527/SAM3-I. |
|
| On the Effectiveness of Textual Prompting with Lightweight Fine-Tuning for SAM3 Remote Sensing Segmentation | 2026-04-12 | ShowRemote sensing (RS) image segmentation is constrained by the limited availability of annotated data and a gap between overhead imagery and natural images used to train foundational models. This motivates effective adaptation under limited supervision. SAM3 concept-driven framework generates masks from textual prompts without requiring task-specific modifications, which may enable this adaptation. We evaluate SAM3 for RS imagery across four target types, comparing textual, geometric, and hybrid prompting strategies, under lightweight fine-tuning scales with increasing supervision, alongside zero-shot inference. Results show that combining semantic and geometric cues yields the highest performance across targets and metrics. Text-only prompting exhibits the lowest performance, with marked score gaps for irregularly shaped targets, reflecting limited semantic alignment between SAM3 textual representations and their overhead appearances. Nevertheless, textual prompting with light fine-tuning offers a practical performance-effort trade-off for geometrically regular and visually salient targets. Across targets, performance improves between zero-shot inference and fine-tuning, followed by diminishing returns as the supervision scale increases. Namely, a modest geometric annotation effort is sufficient for effective adaptation. A persistent gap between Precision and IoU further indicates that under-segmentation and boundary inaccuracies remain prevalent error patterns in RS tasks, particularly for irregular and less prevalent targets. |
|
| Tarot-SAM3: Training-free SAM3 for Any Referring Expression Segmentation | 2026-04-09 | ShowReferring Expression Segmentation (RES) aims to segment image regions described by natural-language expressions, serving as a bridge between vision and language understanding. Existing RES methods, however, rely heavily on large annotated datasets and are limited to either explicit or implicit expressions, hindering their ability to generalize to any referring expression. Recently, the Segment Anything Model 3 (SAM3) has shown impressive robustness in Promptable Concept Segmentation. Nonetheless, applying it to RES remains challenging: (1) SAM3 struggles with longer or implicit expressions; (2) naive coupling of SAM3 with a multimodal large language model (MLLM) makes the final results overly dependent on the MLLM's reasoning capability, without enabling refinement of SAM3's segmentation outputs. To this end, we present Tarot-SAM3, a novel training-free framework that can accurately segment from any referring expression. Specifically, Tarot-SAM3 consists of two key phases. First, the Expression Reasoning Interpreter (ERI) phase introduces reasoning-assisted prompt options to support structured expression parsing and evaluation-aware rephrasing. This transforms arbitrary queries into robust heterogeneous prompts for generating reliable masks with SAM3. Second, the Mask Self-Refining (MSR) phase selects the best mask across prompt types and performs self-refinement by leveraging rich feature relationships from DINOv3 to compare discriminative regions among ERI outputs. It then infers region affiliation to the target, thereby correcting over- and under-segmentation. Extensive experiments demonstrate that Tarot-SAM3 achieves strong performance on both explicit and implicit RES benchmarks, as well as open-world scenarios. Ablation studies further validate the effectiveness of each phase. |
Under review |
| Few-Shot Semantic Segmentation Meets SAM3 | 2026-04-07 | ShowFew-Shot Semantic Segmentation (FSS) focuses on segmenting novel object categories from only a handful of annotated examples. Most existing approaches rely on extensive episodic training to learn transferable representations, which is both computationally demanding and sensitive to distribution shifts. In this work, we revisit FSS from the perspective of modern vision foundation models and explore the potential of Segment Anything Model 3 (SAM3) as a training-free solution. By repurposing its Promptable Concept Segmentation (PCS) capability, we adopt a simple spatial concatenation strategy that places support and query images into a shared canvas, allowing a fully frozen SAM3 to perform segmentation without any fine-tuning or architectural changes. Experiments on PASCAL-$5^i$ and COCO-$20^i$ show that this minimal design already achieves state-of-the-art performance, outperforming many heavily engineered methods. Beyond empirical gains, we uncover that negative prompts can be counterproductive in few-shot settings, where they often weaken target representations and lead to prediction collapse despite their intended role in suppressing distractors. These findings suggest that strong cross-image reasoning can emerge from simple spatial formulations, while also highlighting limitations in how current foundation models handle conflicting prompt signals. Code at: https://github.com/WongKinYiu/FSS-SAM3 |
14 pages, 3 figures |
| The 1st Winner for 5th PVUW MeViS-Text Challenge: Strong MLLMs Meet SAM3 for Referring Video Object Segmentation | 2026-04-01 | ShowThis report presents our winning solution to the 5th PVUW MeViS-Text Challenge. The track studies referring video object segmentation under motion-centric language expressions, where the model must jointly understand appearance, temporal behavior, and object interactions. To address this problem, we build a fully training-free pipeline that combines strong multimodal large language models with SAM3. Our method contains three stages. First, Gemini-3.1 Pro decomposes each target event into instance-level grounding targets, selects the frame where the target is most clearly visible, and generates a discriminative description. Second, SAM3-agent produces a precise seed mask on the selected frame, and the official SAM3 tracker propagates the mask through the whole video. Third, a refinement stage uses Qwen3.5-Plus and behavior-level verification to correct ambiguous or semantically inconsistent predictions. Without task-specific fine-tuning, our method ranks first on the PVUW 2026 MeViS-Text test set, achieving a Final score of 0.909064 and a J&F score of 0.7897. The code is available at https://github.com/Moujuruo/MeViSv2_Track_Solution_2026. |
1st P...1st Place Solution for the 5th PVUW MeViS-Text Challenge (CVPR 2026 Workshop) |
| CAD-Prompted SAM3: Geometry-Conditioned Instance Segmentation for Industrial Objects | 2026-03-24 | ShowVerbal-prompted segmentation is inherently limited by the expressiveness of natural language and struggles with uncommon, instance-specific, or difficult-to-describe objects: scenarios frequently encountered in manufacturing and 3D printing environments. While image exemplars provide an alternative, they primarily encode appearance cues such as color and texture, which are often unrelated to a part's geometric identity. In industrial settings, a single component may be produced in different materials, finishes, or colors, making appearance-based prompting unreliable. In contrast, such objects are typically defined by precise CAD models that capture their canonical geometry. We propose a CAD-prompted segmentation framework built on SAM3 that uses canonical multi-view renderings of a CAD model as prompt input. The rendered views provide geometry-based conditioning independent of surface appearance. The model is trained using synthetic data generated from mesh renderings in simulation under diverse viewpoints and scene contexts. Our approach enables single-stage, CAD-prompted mask prediction, extending promptable segmentation to objects that cannot be robustly described by language or appearance alone. |
|
| Eye image segmentation using visual and concept prompts with Segment Anything Model 3 (SAM3) | 2026-03-18 | ShowPrevious work has reported that vision foundation models show promising zero-shot performance in eye image segmentation. Here we examine whether the latest iteration of the Segment Anything Model, SAM3, offers better eye image segmentation performance than SAM2, and explore the performance of its new concept (text) prompting mode. Eye image segmentation performance was evaluated using diverse datasets encompassing both high-resolution high-quality videos from a lab environment and the TEyeD dataset consisting of challenging eye videos acquired in the wild. Results show that in most cases SAM3 with either visual or concept prompts did not perform better than SAM2, for both lab and in-the-wild datasets. Since SAM2 not only performed better but was also faster, we conclude that SAM2 remains the best option for eye image segmentation. We provide our adaptation of SAM3's codebase that allows processing videos of arbitrary duration. |
|
| ReMeDI: Refined Memory for Disambiguation of Identities with SAM3 in Surgical Segmentation | 2026-03-08 | ShowAccurate surgical instrument segmentation in endoscopy is crucial for computer-assisted interventions, yet remains challenging due to frequent occlusions, rapid motion, and long-term instrument re-entry. While SAM3 provides a powerful spatio-temporal framework for video object segmentation, its performance in surgical scenes is limited by indiscriminate memory updates, fixed memory capacity, and weak identity recovery after occlusions. We propose ReMeDI-SAM3, a training-free extension of SAM3, that addresses these limitations through three components: (i) relevance-aware memory filtering with a dedicated occlusion-aware memory for storing pre-occlusion frames, (ii) a piecewise interpolation scheme that expands effective memory capacity, and (iii) a feature-based re-identification module with temporal voting for reliable post-occlusion identity disambiguation. Together, these components mitigate error accumulation and enable reliable recovery after occlusions. Evaluations on EndoVis17, EndoVis18 and CholecSeg8k under a zero-shot setting show mcIoU improvements of around 5.8%, 8%, and 2% respectively, over vanilla SAM3, outperforming even prior training-based approaches. |
Under Review |
| Progressive Per-Branch Depth Optimization for DEFOM-Stereo and SAM3 Joint Analysis in UAV Forestry Applications | 2026-02-24 | ShowAccurate per-branch 3D reconstruction is a prerequisite for autonomous UAV-based tree pruning; however, dense disparity maps from modern stereo matchers often remain too noisy for individual branch analysis in complex forest canopies. This paper introduces a progressive pipeline integrating DEFOM-Stereo foundation-model disparity estimation, SAM3 instance segmentation, and multi-stage depth optimization to deliver robust per-branch point clouds. Starting from a naive baseline, we systematically identify and resolve three error families through successive refinements. Mask boundary contamination is first addressed through morphological erosion and subsequently refined via a skeleton-preserving variant to safeguard thin-branch topology. Segmentation inaccuracy is then mitigated using LAB-space Mahalanobis color validation coupled with cross-branch overlap arbitration. Finally, depth noise - the most persistent error source - is initially reduced by outlier removal and median filtering, before being superseded by a robust five-stage scheme comprising MAD global detection, spatial density consensus, local MAD filtering, RGB-guided filtering, and adaptive bilateral filtering. Evaluated on 1920x1080 stereo imagery of Radiata pine (Pinus radiata) acquired with a ZED Mini camera (63 mm baseline) from a UAV in Canterbury, New Zealand, the proposed pipeline reduces the average per-branch depth standard deviation by 82% while retaining edge fidelity. The result is geometrically coherent 3D point clouds suitable for autonomous pruning tool positioning. All code and processed data are publicly released to facilitate further UAV forestry research. |
|
| SAM3-LiteText: An Anatomical Study of the SAM3 Text Encoder for Efficient Vision-Language Segmentation | 2026-02-12 | ShowVision-language segmentation models such as SAM3 enable flexible, prompt-driven visual grounding, but inherit large, general-purpose text encoders originally designed for open-ended language understanding. In practice, segmentation prompts are short, structured, and semantically constrained, leading to substantial over-provisioning in text encoder capacity and persistent computational and memory overhead. In this paper, we perform a large-scale anatomical analysis of text prompting in vision-language segmentation, covering 404,796 real prompts across multiple benchmarks. Our analysis reveals severe redundancy: most context windows are underutilized, vocabulary usage is highly sparse, and text embeddings lie on low-dimensional manifold despite high-dimensional representations. Motivated by these findings, we propose SAM3-LiteText, a lightweight text encoding framework that replaces the original SAM3 text encoder with a compact MobileCLIP student that is optimized by knowledge distillation. Extensive experiments on image and video segmentation benchmarks show that SAM3-LiteText reduces text encoder parameters by up to 88%, substantially reducing static memory footprint, while maintaining segmentation performance comparable to the original model. Code: https://github.com/SimonZeng7108/efficientsam3/tree/sam3_litetext. |
|
| Taming SAM3 in the Wild: A Concept Bank for Open-Vocabulary Segmentation | 2026-02-06 | ShowThe recent introduction of \texttt{SAM3} has revolutionized Open-Vocabulary Segmentation (OVS) through \textit{promptable concept segmentation}, which grounds pixel predictions in flexible concept prompts. However, this reliance on pre-defined concepts makes the model vulnerable: when visual distributions shift (\textit{data drift}) or conditional label distributions evolve (\textit{concept drift}) in the target domain, the alignment between visual evidence and prompts breaks down. In this work, we present \textsc{ConceptBank}, a parameter-free calibration framework to restore this alignment on the fly. Instead of adhering to static prompts, we construct a dataset-specific concept bank from the target statistics. Our approach (\textit{i}) anchors target-domain evidence via class-wise visual prototypes, (\textit{ii}) mines representative supports to suppress outliers under data drift, and (\textit{iii}) fuses candidate concepts to rectify concept drift. We demonstrate that \textsc{ConceptBank} effectively adapts \texttt{SAM3} to distribution drifts, including challenging natural-scene and remote-sensing scenarios, establishing a new baseline for robustness and efficiency in OVS. Code and model are available at https://github.com/pgsmall/ConceptBank. |
|
| Medical SAM3: A Foundation Model for Universal Prompt-Driven Medical Image Segmentation | 2026-01-15 | ShowPromptable segmentation foundation models such as SAM3 have demonstrated strong generalization capabilities through interactive and concept-based prompting. However, their direct applicability to medical image segmentation remains limited by severe domain shifts, the absence of privileged spatial prompts, and the need to reason over complex anatomical and volumetric structures. Here we present Medical SAM3, a foundation model for universal prompt-driven medical image segmentation, obtained by fully fine-tuning SAM3 on large-scale, heterogeneous 2D and 3D medical imaging datasets with paired segmentation masks and text prompts. Through a systematic analysis of vanilla SAM3, we observe that its performance degrades substantially on medical data, with its apparent competitiveness largely relying on strong geometric priors such as ground-truth-derived bounding boxes. These findings motivate full model adaptation beyond prompt engineering alone. By fine-tuning SAM3's model parameters on 33 datasets spanning 10 medical imaging modalities, Medical SAM3 acquires robust domain-specific representations while preserving prompt-driven flexibility. Extensive experiments across organs, imaging modalities, and dimensionalities demonstrate consistent and significant performance gains, particularly in challenging scenarios characterized by semantic ambiguity, complex morphology, and long-range 3D context. Our results establish Medical SAM3 as a universal, text-guided segmentation foundation model for medical imaging and highlight the importance of holistic model adaptation for achieving robust prompt-driven segmentation under severe domain shift. Code and model will be made available at https://github.com/AIM-Research-Lab/Medical-SAM3. |
|
| SAM3-DMS: Decoupled Memory Selection for Multi-target Video Segmentation of SAM3 | 2026-01-14 | ShowSegment Anything 3 (SAM3) has established a powerful foundation that robustly detects, segments, and tracks specified targets in videos. However, in its original implementation, its group-level collective memory selection is suboptimal for complex multi-object scenarios, as it employs a synchronized decision across all concurrent targets conditioned on their average performance, often overlooking individual reliability. To this end, we propose SAM3-DMS, a training-free decoupled strategy that utilizes fine-grained memory selection on individual objects. Experiments demonstrate that our approach achieves robust identity preservation and tracking stability. Notably, our advantage becomes more pronounced with increased target density, establishing a solid foundation for simultaneous multi-target video segmentation in the wild. |
Code:... |
| Generalization vs. Specialization: Evaluating Segment Anything Model (SAM3) Zero-Shot Segmentation Against Fine-Tuned YOLO Detectors | 2025-12-09 | ShowDeep learning has advanced two fundamentally different paradigms for instance segmentation: specialized models optimized through task-specific fine-tuning and generalist foundation models capable of zero-shot segmentation. This work presents a comprehensive comparison between SAM3 (Segment Anything Model, also called SAMv3) operating in zero-shot mode and three variants of Ultralytics YOLO11 (nano, medium, and large) fine-tuned for instance segmentation. The evaluation is conducted on the MinneApple dataset, a dense benchmark comprising 670 orchard images with 28,179 annotated apple instances, enabling rigorous validation of model behavior under high object density and occlusion. Our analysis shows IoU choices can inflate performance gaps by up to 30%. At the appropriate IoU = 0.15 threshold, YOLO models achieve 68.9%, 72.2%, and 71.9% F1, while SAM3 reaches 59.8% in pure zero-shot mode. However, YOLO exhibits steep degradation 48-50 points across IoU ranges whereas SAM3 drops only 4 points, revealing 12 times superior boundary stability of SAM3. This highlights the strength of SAMv3 in mask precision versus specialization in detection completeness of YOLO11. We provide open-source code, evaluation pipelines, and methodological recommendations, contributing to a deeper understanding of when specialized fine-tuned models or generalist foundation models are preferable for dense instance segmentation tasks. This project repository is available on GitHub as https://github.com/Applied-AI-Research-Lab/Segment-Anything-Model-SAM3-Zero-Shot-Segmentation-Against-Fine-Tuned-YOLO-Detectors |
|
| The SAM2-to-SAM3 Gap in the Segment Anything Model Family: Why Prompt-Based Expertise Fails in Concept-Driven Image Segmentation | 2025-12-04 | ShowThis paper investigates the fundamental discontinuity between the latest two Segment Anything Models: SAM2 and SAM3. We explain why the expertise in prompt-based segmentation of SAM2 does not transfer to the multimodal concept-driven paradigm of SAM3. SAM2 operates through spatial prompts points, boxes, and masks yielding purely geometric and temporal segmentation. In contrast, SAM3 introduces a unified vision-language architecture capable of open-vocabulary reasoning, semantic grounding, contrastive alignment, and exemplar-based concept understanding. We structure this analysis through five core components: (1) a Conceptual Break Between Prompt-Based and Concept-Based Segmentation, contrasting spatial prompt semantics of SAM2 with multimodal fusion and text-conditioned mask generation of SAM3; (2) Architectural Divergence, detailing pure vision-temporal design of SAM2 versus integration of vision-language encoders, geometry and exemplar encoders, fusion modules, DETR-style decoders, object queries, and ambiguity-handling via Mixture-of-Experts in SAM3; (3) Dataset and Annotation Differences, contrasting SA-V video masks with multimodal concept-annotated corpora of SAM3; (4) Training and Hyperparameter Distinctions, showing why SAM2 optimization knowledge does not apply to SAM3; and (5) Evaluation, Metrics, and Failure Modes, outlining the transition from geometric IoU metrics to semantic, open-vocabulary evaluation. Together, these analyses establish SAM3 as a new class of segmentation foundation model and chart future directions for the emerging concept-driven segmentation era. |
|
| SAM3-UNet: Simplified Adaptation of Segment Anything Model 3 | 2025-12-01 | ShowIn this paper, we introduce SAM3-UNet, a simplified variant of Segment Anything Model 3 (SAM3), designed to adapt SAM3 for downstream tasks at a low cost. Our SAM3-UNet consists of three components: a SAM3 image encoder, a simple adapter for parameter-efficient fine-tuning, and a lightweight U-Net-style decoder. Preliminary experiments on multiple tasks, such as mirror detection and salient object detection, demonstrate that the proposed SAM3-UNet outperforms the prior SAM2-UNet and other state-of-the-art methods, while requiring less than 6 GB of GPU memory during training with a batch size of 12. The code is publicly available at https://github.com/WZH0120/SAM3-UNet. |
Technical Report |
| SAM3-Adapter: Efficient Adaptation of Segment Anything 3 for Camouflage Object Segmentation, Shadow Detection, and Medical Image Segmentation | 2025-11-24 | ShowThe rapid rise of large-scale foundation models has reshaped the landscape of image segmentation, with models such as Segment Anything achieving unprecedented versatility across diverse vision tasks. However, previous generations-including SAM and its successor-still struggle with fine-grained, low-level segmentation challenges such as camouflaged object detection, medical image segmentation, cell image segmentation, and shadow detection. To address these limitations, we originally proposed SAM-Adapter in 2023, demonstrating substantial gains on these difficult scenarios. With the emergence of Segment Anything 3 (SAM3)-a more efficient and higher-performing evolution with a redesigned architecture and improved training pipeline-we revisit these long-standing challenges. In this work, we present SAM3-Adapter, the first adapter framework tailored for SAM3 that unlocks its full segmentation capability. SAM3-Adapter not only reduces computational overhead but also consistently surpasses both SAM and SAM2-based solutions, establishing new state-of-the-art results across multiple downstream tasks, including medical imaging, camouflaged (concealed) object segmentation, and shadow detection. Built upon the modular and composable design philosophy of the original SAM-Adapter, SAM3-Adapter provides stronger generalizability, richer task adaptability, and significantly improved segmentation precision. Extensive experiments confirm that integrating SAM3 with our adapter yields superior accuracy, robustness, and efficiency compared to all prior SAM-based adaptations. We hope SAM3-Adapter can serve as a foundation for future research and practical segmentation applications. Code, pre-trained models, and data processing pipelines are available. |
| Title | Date | Abstract | Comment |
|---|---|---|---|
| A Controlled Benchmark of Visual State-Space Backbones with Domain-Shift and Boundary Analysis for Remote-Sensing Segmentation | 2026-04-20 | ShowVisual state-space models (SSMs) are increasingly promoted as efficient alternatives to Vision Transformers, yet their practical advantages remain unclear under fair comparison because existing studies rarely isolate encoder effects from decoder and training choices. We present a strictly controlled benchmark of representative visual SSM families, including VMamba, MambaVision, and Spatial-Mamba, for remote-sensing semantic segmentation, in which only the encoder varies across experiments. Evaluated on LoveDA and ISPRS Potsdam under a unified 4-stage feature interface and a fixed lightweight decoder, the benchmark reveals three main findings, intra-family scaling yields only modest gains, cross-domain generalization is strongly asymmetric, and boundary delineation is the dominant failure mode under distribution shift. Although visual SSMs achieve favorable accuracy-efficiency trade-offs relative to the controlled CNN and Transformer baselines considered here, the results suggest that future improvements are more likely to come from robustness-oriented design and boundary-aware decoding than from encoder scaling alone. By isolating encoder behavior under a unified and reproducible protocol, this study establishes a practical reference benchmark for the design and evaluation of future Mamba-based segmentation backbones |
5 pag...5 pages, 3 figures, Accepted for publication at IEEE IGARSS 2026 |
| Towards Realistic Open-Vocabulary Remote Sensing Segmentation: Benchmark and Baseline | 2026-04-17 | ShowOpen-vocabulary remote sensing image segmentation (OVRSIS) remains underexplored due to fragmented datasets, limited training diversity, and the lack of evaluation benchmarks that reflect realistic geospatial application demands. Our previous \textit{OVRSISBenchV1} established an initial cross-dataset evaluation protocol, but its limited scope is insufficient for assessing realistic open-world generalization. To address this issue, we propose \textit{OVRSISBenchV2}, a large-scale and application-oriented benchmark for OVRSIS. We first construct \textbf{OVRSIS95K}, a balanced dataset of about 95K image--mask pairs covering 35 common semantic categories across diverse remote sensing scenes. Built upon OVRSIS95K and 10 downstream datasets, OVRSISBenchV2 contains 170K images and 128 categories, substantially expanding scene diversity, semantic coverage, and evaluation difficulty. Beyond standard open-vocabulary segmentation, it further includes downstream protocols for building extraction, road extraction, and flood detection, thereby better reflecting realistic geospatial application demands and complex deployment scenarios. We also propose \textbf{Pi-Seg}, a baseline for OVRSIS. Pi-Seg improves transferability through a \textbf{positive-incentive noise} mechanism, where learnable and semantically guided perturbations broaden the visual-text feature space during training. Extensive experiments on OVRSISBenchV1, OVRSISBenchV2, and downstream tasks show that Pi-Seg delivers strong and consistent results, particularly on the more challenging OVRSISBenchV2 benchmark. Our results highlight both the importance of realistic benchmark design and the effectiveness of perturbation-based transfer for OVRSIS. The code and datasets are available at \href{https://github.com/LiBingyu01/RSKT-Seg/tree/Pi-Seg}{LiBingyu01/RSKT-Seg/tree/Pi-Seg}. |
|
| On the Effectiveness of Textual Prompting with Lightweight Fine-Tuning for SAM3 Remote Sensing Segmentation | 2026-04-12 | ShowRemote sensing (RS) image segmentation is constrained by the limited availability of annotated data and a gap between overhead imagery and natural images used to train foundational models. This motivates effective adaptation under limited supervision. SAM3 concept-driven framework generates masks from textual prompts without requiring task-specific modifications, which may enable this adaptation. We evaluate SAM3 for RS imagery across four target types, comparing textual, geometric, and hybrid prompting strategies, under lightweight fine-tuning scales with increasing supervision, alongside zero-shot inference. Results show that combining semantic and geometric cues yields the highest performance across targets and metrics. Text-only prompting exhibits the lowest performance, with marked score gaps for irregularly shaped targets, reflecting limited semantic alignment between SAM3 textual representations and their overhead appearances. Nevertheless, textual prompting with light fine-tuning offers a practical performance-effort trade-off for geometrically regular and visually salient targets. Across targets, performance improves between zero-shot inference and fine-tuning, followed by diminishing returns as the supervision scale increases. Namely, a modest geometric annotation effort is sufficient for effective adaptation. A persistent gap between Precision and IoU further indicates that under-segmentation and boundary inaccuracies remain prevalent error patterns in RS tasks, particularly for irregular and less prevalent targets. |
|
| SatReg: Regression-based Neural Architecture Search for Lightweight Satellite Image Segmentation | 2026-04-11 | ShowAs Earth-observation workloads move toward onboard and edge processing, remote-sensing segmentation models must operate under tight latency and energy constraints. We present SatReg, a regression-based hardware-aware tuning framework for lightweight remote-sensing segmentation on edge platforms. Using CM-UNet as the teacher architecture, we reduce the search space to two dominant width-related variables, profile a small set of student models on an NVIDIA Jetson Orin Nano, and fit low-order surrogate models for mIoU, latency, and power. Knowledge distillation is used to efficiently train the sampled students. The learned surrogates enable fast selection of near-optimal architecture settings for deployment targets without exhaustive search. Results show that the selected variables affect task accuracy and hardware cost differently, making reduced-space regression a practical strategy for adapting hybrid CNN-Mamba segmentation models to future space-edge systems. |
|
| ProtoFlow: Mitigating Forgetting in Class-Incremental Remote Sensing Segmentation via Low-Curvature Prototype Flow | 2026-04-03 | ShowRemote sensing segmentation in real deployment is inherently continual: new semantic categories emerge, and acquisition conditions shift across seasons, cities, and sensors. Despite recent progress, many incremental approaches still treat training steps as isolated updates, which leaves representation drift and forgetting insufficiently controlled. We present ProtoFlow, a time-aware prototype dynamics framework that models class prototypes as trajectories and learns their evolution with an explicit temporal vector field. By jointly enforcing low-curvature motion and inter-class separation, ProtoFlow stabilizes prototype geometry throughout incremental learning. Experiments on standard class- and domain-incremental remote sensing benchmarks show consistent gains over strong baselines, including up to 1.5-2.0 points improvement in mIoUall, together with reduced forgetting. These results suggest that explicitly modeling temporal prototype evolution is a practical and interpretable strategy for robust continual remote sensing segmentation. |
|
| Decouple and Rectify: Semantics-Preserving Structural Enhancement for Open-Vocabulary Remote Sensing Segmentation | 2026-04-02 | ShowOpen-vocabulary semantic segmentation in the remote sensing (RS) field requires both language-aligned recognition and fine-grained spatial delineation. Although CLIP offers robust semantic generalization, its global-aligned visual representations inherently struggle to capture structural details. Recent methods attempt to compensate for this by introducing RS-pretrained DINO features. However, these methods treat CLIP representations as a monolithic semantic space and cannot localize where structural enhancement is required, failing to effectively delineate boundaries while risking the disruption of CLIP's semantic integrity. To address this limitation, we propose DR-Seg, a novel decouple-and-rectify framework in this paper. Our method is motivated by the key observation that CLIP feature channels exhibit distinct functional heterogeneity rather than forming a uniform semantic space. Building on this insight, DR-Seg decouples CLIP features into semantics-dominated and structure-dominated subspaces, enabling targeted structural enhancement by DINO without distorting language-aligned semantics. Subsequently, a prior-driven graph rectification module injects high-fidelity structural priors under DINO guidance to form a refined branch, while an uncertainty-guided adaptive fusion module dynamically integrates this refined branch with the original CLIP branch for final prediction. Comprehensive experiments across eight benchmarks demonstrate that DR-Seg establishes a new state-of-the-art. |
|
| ConInfer: Context-Aware Inference for Training-Free Open-Vocabulary Remote Sensing Segmentation | 2026-03-31 | ShowTraining-free open-vocabulary remote sensing segmentation (OVRSS), empowered by vision-language models, has emerged as a promising paradigm for achieving category-agnostic semantic understanding in remote sensing imagery. Existing approaches mainly focus on enhancing feature representations or mitigating modality discrepancies to improve patch-level prediction accuracy. However, such independent prediction schemes are fundamentally misaligned with the intrinsic characteristics of remote sensing data. In real-world applications, remote sensing scenes are typically large-scale and exhibit strong spatial as well as semantic correlations, making isolated patch-wise predictions insufficient for accurate segmentation. To address this limitation, we propose ConInfer, a context-aware inference framework for OVRSS that performs joint prediction across multiple spatial units while explicitly modeling their inter-unit semantic dependencies. By incorporating global contextual cues, our method significantly enhances segmentation consistency, robustness, and generalization in complex remote sensing environments. Extensive experiments on multiple benchmark datasets demonstrate that our approach consistently surpasses state-of-the-art per-pixel VLM-based baselines such as SegEarth-OV, achieving average improvements of 2.80% and 6.13% on open-vocabulary semantic segmentation and object extraction tasks, respectively. The implementation code is available at: https://github.com/Dog-Yang/ConInfer |
|
| Transferring Physical Priors into Remote Sensing Segmentation via Large Language Models | 2026-03-29 | ShowSemantic segmentation of remote sensing imagery is fundamental to Earth observation. Achieving accurate results requires integrating not only optical images but also physical variables such as the Digital Elevation Model (DEM), Synthetic Aperture Radar (SAR) and Normalized Difference Vegetation Index (NDVI). Recent foundation models (FMs) leverage pre-training to exploit these variables but still depend on spatially aligned data and costly retraining when involving new sensors. To overcome these limitations, we introduce a novel paradigm for integrating domain-specific physical priors into segmentation models. We first construct a Physical-Centric Knowledge Graph (PCKG) by prompting large language models to extract physical priors from 1,763 vocabularies, and use it to build a heterogeneous, spatial-aligned dataset, Phy-Sky-SA. Building on this foundation, we develop PriorSeg, a physics-aware residual refinement model trained with a joint visual-physical strategy that incorporates a novel physics-consistency loss. Experiments on heterogeneous settings demonstrate that PriorSeg improves segmentation accuracy and physical plausibility without retraining the FMs. Ablation studies verify the effectiveness of the Phy-Sky-SA dataset, the PCKG, and the physics-consistency loss. |
|
| Cross-Modal Urban Sensing: Evaluating Sound-Vision Alignment Across Street-Level and Aerial Imagery | 2026-03-27 | ShowEnvironmental soundscapes convey substantial ecological and social information regarding urban environments; however, their potential remains largely untapped in large-scale geographic analysis. In this study, we investigate the extent to which urban sounds correspond with visual scenes by comparing various visual representation strategies in capturing acoustic semantics. We employ a multimodal approach that integrates geo-referenced sound recordings with both street-level and remote sensing imagery across three major global cities: London, New York, and Tokyo. Utilizing the AST model for audio, along with CLIP and RemoteCLIP for imagery, as well as CLIPSeg and Seg-Earth OV for semantic segmentation, we extract embeddings and class-level features to evaluate cross-modal similarity. The results indicate that street view embeddings demonstrate stronger alignment with environmental sounds compared to segmentation outputs, whereas remote sensing segmentation is more effective in interpreting ecological categories through a Biophony--Geophony--Anthrophony (BGA) framework. These findings imply that embedding-based models offer superior semantic alignment, while segmentation-based methods provide interpretable links between visual structure and acoustic ecology. This work advances the burgeoning field of multimodal urban sensing by offering novel perspectives for incorporating sound into geospatial analysis. |
18 pages, 13 figures |
| LEMMA: Laplacian pyramids for Efficient Marine SeMAntic Segmentation | 2026-03-26 | ShowSemantic segmentation in marine environments is crucial for the autonomous navigation of unmanned surface vessels (USVs) and coastal Earth Observation events such as oil spills. However, existing methods, often relying on deep CNNs and transformer-based architectures, face challenges in deployment due to their high computational costs and resource-intensive nature. These limitations hinder the practicality of real-time, low-cost applications in real-world marine settings. To address this, we propose LEMMA, a lightweight semantic segmentation model designed specifically for accurate remote sensing segmentation under resource constraints. The proposed architecture leverages Laplacian Pyramids to enhance edge recognition, a critical component for effective feature extraction in complex marine environments for disaster response, environmental surveillance, and coastal monitoring. By integrating edge information early in the feature extraction process, LEMMA eliminates the need for computationally expensive feature map computations in deeper network layers, drastically reducing model size, complexity and inference time. LEMMA demonstrates state-of-the-art performance across datasets captured from diverse platforms while reducing trainable parameters and computational requirements by up to 71x, GFLOPs by up to 88.5%, and inference time by up to 84.65%, as compared to existing models. Experimental results highlight its effectiveness and real-world applicability, including 93.42% IoU on the Oil Spill dataset and 98.97% mIoU on Mastr1325. |
Accep...Accepted at the MaCVi Workshop, CVPR 2026 |
| GeoSeg: Training-Free Reasoning-Driven Segmentation in Remote Sensing Imagery | 2026-03-04 | ShowRecent advances in MLLMs are reframing segmentation from fixed-category prediction to instruction-grounded localization. While reasoning based segmentation has progressed rapidly in natural scenes, remote sensing lacks a generalizable solution due to the prohibitive cost of reasoning-oriented data and domain-specific challenges like overhead viewpoints. We present GeoSeg, a zero-shot, training-free framework that bypasses the supervision bottleneck for reasoning-driven remote sensing segmentation. GeoSeg couples MLLM reasoning with precise localization via: (i) bias-aware coordinate refinement to correct systematic grounding shifts and (ii) a dual-route prompting mechanism to fuse semantic intent with fine-grained spatial cues. We also introduce GeoSeg-Bench, a diagnostic benchmark of 810 image--query pairs with hierarchical difficulty levels. Experiments show that GeoSeg consistently outperforms all baselines, with extensive ablations confirming the effectiveness and necessity of each component. |
|
| Enabling Training-Free Text-Based Remote Sensing Segmentation | 2026-02-19 | ShowRecent advances in Vision Language Models (VLMs) and Vision Foundation Models (VFMs) have opened new opportunities for zero-shot text-guided segmentation of remote sensing imagery. However, most existing approaches still rely on additional trainable components, limiting their generalisation and practical applicability. In this work, we investigate to what extent text-based remote sensing segmentation can be achieved without additional training, by relying solely on existing foundation models. We propose a simple yet effective approach that integrates contrastive and generative VLMs with the Segment Anything Model (SAM), enabling a fully training-free or lightweight LoRA-tuned pipeline. Our contrastive approach employs CLIP as mask selector for SAM's grid-based proposals, achieving state-of-the-art open-vocabulary semantic segmentation (OVSS) in a completely zero-shot setting. In parallel, our generative approach enables reasoning and referring segmentation by generating click prompts for SAM using GPT-5 in a zero-shot setting and a LoRA-tuned Qwen-VL model, with the latter yielding the best results. Extensive experiments across 19 remote sensing benchmarks, including open-vocabulary, referring, and reasoning-based tasks, demonstrate the strong capabilities of our approach. Code will be released at https://github.com/josesosajs/trainfree-rs-segmentation. |
|
| DIS2: Disentanglement Meets Distillation with Classwise Attention for Robust Remote Sensing Segmentation under Missing Modalities | 2026-01-20 | ShowThe efficacy of multimodal learning in remote sensing (RS) is severely undermined by missing modalities. The challenge is exacerbated by the RS highly heterogeneous data and huge scale variation. Consequently, paradigms proven effective in other domains often fail when confronted with these unique data characteristics. Conventional disentanglement learning, which relies on significant feature overlap between modalities (modality-invariant), is insufficient for this heterogeneity. Similarly, knowledge distillation becomes an ill-posed mimicry task where a student fails to focus on the necessary compensatory knowledge, leaving the semantic gap unaddressed. Our work is therefore built upon three pillars uniquely designed for RS: (1) principled missing information compensation, (2) class-specific modality contribution, and (3) multi-resolution feature importance. We propose a novel method DIS2, a new paradigm shifting from modality-shared feature dependence and untargeted imitation to active, guided missing features compensation. Its core novelty lies in a reformulated synergy between disentanglement learning and knowledge distillation, termed DLKD. Compensatory features are explicitly captured which, when fused with the features of the available modality, approximate the ideal fused representation of the full-modality case. To address the class-specific challenge, our Classwise Feature Learning Module (CFLM) adaptively learn discriminative evidence for each target depending on signal availability. Both DLKD and CFLM are supported by a hierarchical hybrid fusion (HF) structure using features across resolutions to strengthen prediction. Extensive experiments validate that our proposed approach significantly outperforms state-of-the-art methods across benchmarks. |
Accep...Accepted to WACV 2026 - Computer Vision for Earth Observation Workshop |
| Toward Stable Semi-Supervised Remote Sensing Segmentation via Co-Guidance and Co-Fusion | 2026-01-12 | ShowSemi-supervised remote sensing (RS) image semantic segmentation offers a promising solution to alleviate the burden of exhaustive annotation, yet it fundamentally struggles with pseudo-label drift, a phenomenon where confirmation bias leads to the accumulation of errors during training. In this work, we propose Co2S, a stable semi-supervised RS segmentation framework that synergistically fuses priors from vision-language models and self-supervised models. Specifically, we construct a heterogeneous dual-student architecture comprising two distinct ViT-based vision foundation models initialized with pretrained CLIP and DINOv3 to mitigate error accumulation and pseudo-label drift. To effectively incorporate these distinct priors, an explicit-implicit semantic co-guidance mechanism is introduced that utilizes text embeddings and learnable queries to provide explicit and implicit class-level guidance, respectively, thereby jointly enhancing semantic consistency. Furthermore, a global-local feature collaborative fusion strategy is developed to effectively fuse the global contextual information captured by CLIP with the local details produced by DINOv3, enabling the model to generate highly precise segmentation results. Extensive experiments on six popular datasets demonstrate the superiority of the proposed method, which consistently achieves leading performance across various partition protocols and diverse scenarios. Project page is available at https://xavierjiezou.github.io/Co2S/. |
12 pa...12 pages, 5 figures, 9 tables |
| SAM2-ELNet: Label Enhancement and Automatic Annotation for Remote Sensing Segmentation | 2025-09-21 | ShowRemote sensing image segmentation is crucial for environmental monitoring, disaster assessment, and resource management, but its performance largely depends on the quality of the dataset. Although several high-quality datasets are broadly accessible, data scarcity remains for specialized tasks like marine oil spill segmentation. Such tasks still rely on manual annotation, which is both time-consuming and influenced by subjective human factors. The segment anything model 2 (SAM2) has strong potential as an automatic annotation framework but struggles to perform effectively on heterogeneous, low-contrast remote sensing imagery. To address these challenges, we introduce a novel label enhancement and automatic annotation framework, termed SAM2-ELNet (Enhancement and Labeling Network). Specifically, we employ the frozen Hiera backbone from the pretrained SAM2 as the encoder, while fine-tuning the adapter and decoder for different remote sensing tasks. In addition, the proposed framework includes a label quality evaluator for filtering, ensuring the reliability of the generated labels. We design a series of experiments targeting resource-limited remote sensing tasks and evaluate our method on two datasets: the Deep-SAR Oil Spill (SOS) dataset with Synthetic Aperture Radar (SAR) imagery, and the CHN6-CUG Road dataset with Very High Resolution (VHR) optical imagery. The proposed framework can enhance coarse annotations and generate reliable training data under resource-limited conditions. Fine-tuned on only 30% of the training data, it generates automatically labeled data. A model trained solely on these achieves slightly lower performance than using the full original annotations, while greatly reducing labeling costs and offering a practical solution for large-scale remote sensing interpretation. |
publi...published in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| TASAM: Terrain-and-Aware Segment Anything Model for Temporal-Scale Remote Sensing Segmentation | 2025-09-19 | ShowSegment Anything Model (SAM) has demonstrated impressive zero-shot segmentation capabilities across natural image domains, but it struggles to generalize to the unique challenges of remote sensing data, such as complex terrain, multi-scale objects, and temporal dynamics. In this paper, we introduce TASAM, a terrain and temporally-aware extension of SAM designed specifically for high-resolution remote sensing image segmentation. TASAM integrates three lightweight yet effective modules: a terrain-aware adapter that injects elevation priors, a temporal prompt generator that captures land-cover changes over time, and a multi-scale fusion strategy that enhances fine-grained object delineation. Without retraining the SAM backbone, our approach achieves substantial performance gains across three remote sensing benchmarks-LoveDA, iSAID, and WHU-CD-outperforming both zero-shot SAM and task-specific models with minimal computational overhead. Our results highlight the value of domain-adaptive augmentation for foundation models and offer a scalable path toward more robust geospatial segmentation. |
|
| Baltimore Atlas: FreqWeaver Adapter for Semi-supervised Ultra-high Spatial Resolution Land Cover Classification | 2025-08-23 | ShowUltra-high Spatial Resolution (UHSR) Land Cover Classification is increasingly important for urban analysis, enabling fine-scale planning, ecological monitoring, and infrastructure management. It identifies land cover types on sub-meter remote sensing imagery, capturing details such as building outlines, road networks, and distinct boundaries. However, most existing methods focus on 1 m imagery and rely heavily on large-scale annotations, while UHSR data remain scarce and difficult to annotate, limiting practical applicability. To address these challenges, we introduce Baltimore Atlas, a UHSR land cover classification framework that reduces reliance on large-scale training data and delivers high-accuracy results. Baltimore Atlas builds on three key ideas: (1) Baltimore Atlas Dataset, a 0.3 m resolution dataset based on aerial imagery of Baltimore City; (2) FreqWeaver Adapter, a parameter-efficient adapter that transfers SAM2 to this domain, leveraging foundation model knowledge to reduce training data needs while enabling fine-grained detail and structural modeling; (3) Uncertainty-Aware Teacher Student Framework, a semi-supervised framework that exploits unlabeled data to further reduce training dependence and improve generalization across diverse scenes. Using only 5.96% of total model parameters, our approach achieves a 1.78% IoU improvement over existing parameter-efficient tuning strategies and a 3.44% IoU gain compared to state-of-the-art high-resolution remote sensing segmentation methods on the Baltimore Atlas Dataset. |
|
| ViRefSAM: Visual Reference-Guided Segment Anything Model for Remote Sensing Segmentation | 2025-07-03 | ShowThe Segment Anything Model (SAM), with its prompt-driven paradigm, exhibits strong generalization in generic segmentation tasks. However, applying SAM to remote sensing (RS) images still faces two major challenges. First, manually constructing precise prompts for each image (e.g., points or boxes) is labor-intensive and inefficient, especially in RS scenarios with dense small objects or spatially fragmented distributions. Second, SAM lacks domain adaptability, as it is pre-trained primarily on natural images and struggles to capture RS-specific semantics and spatial characteristics, especially when segmenting novel or unseen classes. To address these issues, inspired by few-shot learning, we propose ViRefSAM, a novel framework that guides SAM utilizing only a few annotated reference images that contain class-specific objects. Without requiring manual prompts, ViRefSAM enables automatic segmentation of class-consistent objects across RS images. Specifically, ViRefSAM introduces two key components while keeping SAM's original architecture intact: (1) a Visual Contextual Prompt Encoder that extracts class-specific semantic clues from reference images and generates object-aware prompts via contextual interaction with target images; and (2) a Dynamic Target Alignment Adapter, integrated into SAM's image encoder, which mitigates the domain gap by injecting class-specific semantics into target image features, enabling SAM to dynamically focus on task-relevant regions. Extensive experiments on three few-shot segmentation benchmarks, including iSAID-5$^i$, LoveDA-2$^i$, and COCO-20$^i$, demonstrate that ViRefSAM enables accurate and automatic segmentation of unseen classes by leveraging only a few reference images and consistently outperforms existing few-shot segmentation methods across diverse datasets. |
|
| RS-MTDF: Multi-Teacher Distillation and Fusion for Remote Sensing Semi-Supervised Semantic Segmentation | 2025-06-11 | ShowSemantic segmentation in remote sensing images is crucial for various applications, yet its performance is heavily reliant on large-scale, high-quality pixel-wise annotations, which are notoriously expensive and time-consuming to acquire. Semi-supervised semantic segmentation (SSS) offers a promising alternative to mitigate this data dependency. However, existing SSS methods often struggle with the inherent distribution mismatch between limited labeled data and abundant unlabeled data, leading to suboptimal generalization. To alleviate this issue, we attempt to introduce the Vision Foundation Models (VFMs) pre-trained on vast and diverse datasets into the SSS task since VFMs possess robust generalization capabilities that can effectively bridge this distribution gap and provide strong semantic priors for SSS. Inspired by this, we introduce RS-MTDF (Multi-Teacher Distillation and Fusion), a novel framework that leverages the powerful semantic knowledge embedded in VFMs to guide semi-supervised learning in remote sensing. Specifically, RS-MTDF employs multiple frozen VFMs (e.g., DINOv2 and CLIP) as expert teachers, utilizing feature-level distillation to align student features with their robust representations. To further enhance discriminative power, the distilled knowledge is seamlessly fused into the student decoder. Extensive experiments on three challenging remote sensing datasets demonstrate that RS-MTDF consistently achieves state-of-the-art performance. Notably, our method outperforms existing approaches across various label ratios on LoveDA and secures the highest IoU in the majority of semantic categories. These results underscore the efficacy of multi-teacher VFM guidance in significantly enhancing both generalization and semantic understanding for remote sensing segmentation. Ablation studies further validate the contribution of each proposed module. |
|
| F2Net: A Frequency-Fused Network for Ultra-High Resolution Remote Sensing Segmentation | 2025-06-09 | ShowSemantic segmentation of ultra-high-resolution (UHR) remote sensing imagery is critical for applications like environmental monitoring and urban planning but faces computational and optimization challenges. Conventional methods either lose fine details through downsampling or fragment global context via patch processing. While multi-branch networks address this trade-off, they suffer from computational inefficiency and conflicting gradient dynamics during training. We propose F2Net, a frequency-aware framework that decomposes UHR images into high- and low-frequency components for specialized processing. The high-frequency branch preserves full-resolution structural details, while the low-frequency branch processes downsampled inputs through dual sub-branches capturing short- and long-range dependencies. A Hybrid-Frequency Fusion module integrates these observations, guided by two novel objectives: Cross-Frequency Alignment Loss ensures semantic consistency between frequency components, and Cross-Frequency Balance Loss regulates gradient magnitudes across branches to stabilize training. Evaluated on DeepGlobe and Inria Aerial benchmarks, F2Net achieves state-of-the-art performance with mIoU of 80.22 and 83.39, respectively. Our code will be publicly available. |
|
| AerOSeg: Harnessing SAM for Open-Vocabulary Segmentation in Remote Sensing Images | 2025-04-12 | ShowImage segmentation beyond predefined categories is a key challenge in remote sensing, where novel and unseen classes often emerge during inference. Open-vocabulary image Segmentation addresses these generalization issues in traditional supervised segmentation models while reducing reliance on extensive per-pixel annotations, which are both expensive and labor-intensive to obtain. Most Open-Vocabulary Segmentation (OVS) methods are designed for natural images but struggle with remote sensing data due to scale variations, orientation changes, and complex scene compositions. This necessitates the development of OVS approaches specifically tailored for remote sensing. In this context, we propose AerOSeg, a novel OVS approach for remote sensing data. First, we compute robust image-text correlation features using multiple rotated versions of the input image and domain-specific prompts. These features are then refined through spatial and class refinement blocks. Inspired by the success of the Segment Anything Model (SAM) in diverse domains, we leverage SAM features to guide the spatial refinement of correlation features. Additionally, we introduce a semantic back-projection module and loss to ensure the seamless propagation of SAM's semantic information throughout the segmentation pipeline. Finally, we enhance the refined correlation features using a multi-scale attention-aware decoder to produce the final segmentation map. We validate our SAM-guided Open-Vocabulary Remote Sensing Segmentation model on three benchmark remote sensing datasets: iSAID, DLRSD, and OpenEarthMap. Our model outperforms state-of-the-art open-vocabulary segmentation methods, achieving an average improvement of 2.54 h-mIoU. |
Accep...Accepted at EarthVision workshop, CVPR 2025 |
| Center-guided Classifier for Semantic Segmentation of Remote Sensing Images | 2025-03-21 | ShowCompared with natural images, remote sensing images (RSIs) have the unique characteristic. i.e., larger intraclass variance, which makes semantic segmentation for remote sensing images more challenging. Moreover, existing semantic segmentation models for remote sensing images usually employ a vanilla softmax classifier, which has three drawbacks: (1) non-direct supervision for the pixel representations during training; (2) inadequate modeling ability of parametric softmax classifiers under large intraclass variance; and (3) opaque process of classification decision. In this paper, we propose a novel classifier (called CenterSeg) customized for RSI semantic segmentation, which solves the abovementioned problems with multiple prototypes, direct supervision under Grassmann manifold, and interpretability strategy. Specifically, for each class, our CenterSeg obtains local class centers by aggregating corresponding pixel features based on ground-truth masks, and generates multiple prototypes through hard attention assignment and momentum updating. In addition, we introduce the Grassmann manifold and constrain the joint embedding space of pixel features and prototypes based on two additional regularization terms. Especially, during the inference, CenterSeg can further provide interpretability to the model by restricting the prototype as a sample of the training set. Experimental results on three remote sensing segmentation datasets validate the effectiveness of the model. Besides the superior performance, CenterSeg has the advantages of simplicity, lightweight, compatibility, and interpretability. Code is available at https://github.com/xwmaxwma/rssegmentation. |
|
| Joint-Optimized Unsupervised Adversarial Domain Adaptation in Remote Sensing Segmentation with Prompted Foundation Model | 2024-11-18 | ShowUnsupervised Domain Adaptation for Remote Sensing Semantic Segmentation (UDA-RSSeg) addresses the challenge of adapting a model trained on source domain data to target domain samples, thereby minimizing the need for annotated data across diverse remote sensing scenes. This task presents two principal challenges: (1) severe inconsistencies in feature representation across different remote sensing domains, and (2) a domain gap that emerges due to the representation bias of source domain patterns when translating features to predictive logits. To tackle these issues, we propose a joint-optimized adversarial network incorporating the "Segment Anything Model (SAM) (SAM-JOANet)" for UDA-RSSeg. Our approach integrates SAM to leverage its robust generalized representation capabilities, thereby alleviating feature inconsistencies. We introduce a finetuning decoder designed to convert SAM-Encoder features into predictive logits. Additionally, a feature-level adversarial-based prompted segmentor is employed to generate class-agnostic maps, which guide the finetuning decoder's feature representations. The network is optimized end-to-end, combining the prompted segmentor and the finetuning decoder. Extensive evaluations on benchmark datasets, including ISPRS (Potsdam/Vaihingen) and CITY-OSM (Paris/Chicago), demonstrate the effectiveness of our method. The results, supported by visualization and analysis, confirm the method's interpretability and robustness. The code of this paper is available at https://github.com/CV-ShuchangLyu/SAM-JOANet. |
12 pa...12 pages,6 figures, 6 tables |
| Graph Information Bottleneck for Remote Sensing Segmentation | 2024-08-31 | ShowRemote sensing segmentation has a wide range of applications in environmental protection, and urban change detection, etc. Despite the success of deep learning-based remote sensing segmentation methods (e.g., CNN and Transformer), they are not flexible enough to model irregular objects. In addition, existing graph contrastive learning methods usually adopt the way of maximizing mutual information to keep the node representations consistent between different graph views, which may cause the model to learn task-independent redundant information. To tackle the above problems, this paper treats images as graph structures and introduces a simple contrastive vision GNN (SC-ViG) architecture for remote sensing segmentation. Specifically, we construct a node-masked and edge-masked graph view to obtain an optimal graph structure representation, which can adaptively learn whether to mask nodes and edges. Furthermore, this paper innovatively introduces information bottleneck theory into graph contrastive learning to maximize task-related information while minimizing task-independent redundant information. Finally, we replace the convolutional module in UNet with the SC-ViG module to complete the segmentation and classification tasks of remote sensing images. Extensive experiments on publicly available real datasets demonstrate that our method outperforms state-of-the-art remote sensing image segmentation methods. |
13 pages, 6 figures |
| Hi-ResNet: Edge Detail Enhancement for High-Resolution Remote Sensing Segmentation | 2024-08-15 | ShowHigh-resolution remote sensing (HRS) semantic segmentation extracts key objects from high-resolution coverage areas. However, objects of the same category within HRS images generally show significant differences in scale and shape across diverse geographical environments, making it difficult to fit the data distribution. Additionally, a complex background environment causes similar appearances of objects of different categories, which precipitates a substantial number of objects into misclassification as background. These issues make existing learning algorithms sub-optimal. In this work, we solve the above-mentioned problems by proposing a High-resolution remote sensing network (Hi-ResNet) with efficient network structure designs, which consists of a funnel module, a multi-branch module with stacks of information aggregation (IA) blocks, and a feature refinement module, sequentially, and Class-agnostic Edge Aware (CEA) loss. Specifically, we propose a funnel module to downsample, which reduces the computational cost, and extract high-resolution semantic information from the initial input image. Secondly, we downsample the processed feature images into multi-resolution branches incrementally to capture image features at different scales and apply IA blocks, which capture key latent information by leveraging attention mechanisms, for effective feature aggregation, distinguishing image features of the same class with variant scales and shapes. Finally, our feature refinement module integrate the CEA loss function, which disambiguates inter-class objects with similar shapes and increases the data distribution distance for correct predictions. With effective pre-training strategies, we demonstrated the superiority of Hi-ResNet over state-of-the-art methods on three HRS segmentation benchmarks. |
|
| ALPS: An Auto-Labeling and Pre-training Scheme for Remote Sensing Segmentation With Segment Anything Model | 2024-06-16 | ShowIn the fast-growing field of Remote Sensing (RS) image analysis, the gap between massive unlabeled datasets and the ability to fully utilize these datasets for advanced RS analytics presents a significant challenge. To fill the gap, our work introduces an innovative auto-labeling framework named ALPS (Automatic Labeling for Pre-training in Segmentation), leveraging the Segment Anything Model (SAM) to predict precise pseudo-labels for RS images without necessitating prior annotations or additional prompts. The proposed pipeline significantly reduces the labor and resource demands traditionally associated with annotating RS datasets. By constructing two comprehensive pseudo-labeled RS datasets via ALPS for pre-training purposes, our approach enhances the performance of downstream tasks across various benchmarks, including iSAID and ISPRS Potsdam. Experiments demonstrate the effectiveness of our framework, showcasing its ability to generalize well across multiple tasks even under the scarcity of extensively annotated datasets, offering a scalable solution to automatic segmentation and annotation challenges in the field. In addition, the proposed a pipeline is flexible and can be applied to medical image segmentation, remarkably boosting the performance. Note that ALPS utilizes pre-trained SAM to semi-automatically annotate RS images without additional manual annotations. Though every component in the pipeline has bee well explored, integrating clustering algorithms with SAM and novel pseudo-label alignment significantly enhances RS segmentation, as an off-the-shelf tool for pre-training data preparation. Our source code is available at: https://github.com/StriveZs/ALPS. |
|
| Embedding Generalized Semantic Knowledge into Few-Shot Remote Sensing Segmentation | 2024-05-22 | ShowFew-shot segmentation (FSS) for remote sensing (RS) imagery leverages supporting information from limited annotated samples to achieve query segmentation of novel classes. Previous efforts are dedicated to mining segmentation-guiding visual cues from a constrained set of support samples. However, they still struggle to address the pronounced intra-class differences in RS images, as sparse visual cues make it challenging to establish robust class-specific representations. In this paper, we propose a holistic semantic embedding (HSE) approach that effectively harnesses general semantic knowledge, i.e., class description (CD) embeddings.Instead of the naive combination of CD embeddings and visual features for segmentation decoding, we investigate embedding the general semantic knowledge during the feature extraction stage.Specifically, in HSE, a spatial dense interaction module allows the interaction of visual support features with CD embeddings along the spatial dimension via self-attention.Furthermore, a global content modulation module efficiently augments the global information of the target category in both support and query features, thanks to the transformative fusion of visual features and CD embeddings.These two components holistically synergize general CD embeddings and visual cues, constructing a robust class-specific representation.Through extensive experiments on the standard FSS benchmark, the proposed HSE approach demonstrates superior performance compared to peer work, setting a new state-of-the-art. |
|
| SFFNet: A Wavelet-Based Spatial and Frequency Domain Fusion Network for Remote Sensing Segmentation | 2024-05-03 | ShowIn order to fully utilize spatial information for segmentation and address the challenge of handling areas with significant grayscale variations in remote sensing segmentation, we propose the SFFNet (Spatial and Frequency Domain Fusion Network) framework. This framework employs a two-stage network design: the first stage extracts features using spatial methods to obtain features with sufficient spatial details and semantic information; the second stage maps these features in both spatial and frequency domains. In the frequency domain mapping, we introduce the Wavelet Transform Feature Decomposer (WTFD) structure, which decomposes features into low-frequency and high-frequency components using the Haar wavelet transform and integrates them with spatial features. To bridge the semantic gap between frequency and spatial features, and facilitate significant feature selection to promote the combination of features from different representation domains, we design the Multiscale Dual-Representation Alignment Filter (MDAF). This structure utilizes multiscale convolutions and dual-cross attentions. Comprehensive experimental results demonstrate that, compared to existing methods, SFFNet achieves superior performance in terms of mIoU, reaching 84.80% and 87.73% respectively.The code is located at https://github.com/yysdck/SFFNet. |
|
| RSAM-Seg: A SAM-based Approach with Prior Knowledge Integration for Remote Sensing Image Semantic Segmentation | 2024-02-29 | ShowThe development of high-resolution remote sensing satellites has provided great convenience for research work related to remote sensing. Segmentation and extraction of specific targets are essential tasks when facing the vast and complex remote sensing images. Recently, the introduction of Segment Anything Model (SAM) provides a universal pre-training model for image segmentation tasks. While the direct application of SAM to remote sensing image segmentation tasks does not yield satisfactory results, we propose RSAM-Seg, which stands for Remote Sensing SAM with Semantic Segmentation, as a tailored modification of SAM for the remote sensing field and eliminates the need for manual intervention to provide prompts. Adapter-Scale, a set of supplementary scaling modules, are proposed in the multi-head attention blocks of the encoder part of SAM. Furthermore, Adapter-Feature are inserted between the Vision Transformer (ViT) blocks. These modules aim to incorporate high-frequency image information and image embedding features to generate image-informed prompts. Experiments are conducted on four distinct remote sensing scenarios, encompassing cloud detection, field monitoring, building detection and road mapping tasks . The experimental results not only showcase the improvement over the original SAM and U-Net across cloud, buildings, fields and roads scenarios, but also highlight the capacity of RSAM-Seg to discern absent areas within the ground truth of certain datasets, affirming its potential as an auxiliary annotation method. In addition, the performance in few-shot scenarios is commendable, underscores its potential in dealing with limited datasets. |
12 pages, 11 figures |
| A Framework of Full-Process Generation Design for Park Green Spaces Based on Remote Sensing Segmentation-GAN-Diffusion | 2023-12-17 | ShowThe development of generative design driven by artificial intelligence algorithms is speedy. There are two research gaps in the current research: 1) Most studies only focus on the relationship between design elements and pay little attention to the external information of the site; 2) GAN and other traditional generative algorithms generate results with low resolution and insufficient details. To address these two problems, we integrate GAN, Stable diffusion multimodal large-scale image pre-training model to construct a full-process park generative design method: 1) First, construct a high-precision remote sensing object extraction system for automated extraction of urban environmental information; 2) Secondly, use GAN to construct a park design generation system based on the external environment, which can quickly infer and generate design schemes from urban environmental information; 3) Finally, introduce Stable Diffusion to optimize the design plan, fill in details, and expand the resolution of the plan by 64 times. This method can achieve a fully unmanned design automation workflow. The research results show that: 1) The relationship between the inside and outside of the site will affect the algorithm generation results. 2) Compared with traditional GAN algorithms, Stable diffusion significantly improve the information richness of the generated results. |
|
| SAMRS: Scaling-up Remote Sensing Segmentation Dataset with Segment Anything Model | 2023-10-13 | ShowThe success of the Segment Anything Model (SAM) demonstrates the significance of data-centric machine learning. However, due to the difficulties and high costs associated with annotating Remote Sensing (RS) images, a large amount of valuable RS data remains unlabeled, particularly at the pixel level. In this study, we leverage SAM and existing RS object detection datasets to develop an efficient pipeline for generating a large-scale RS segmentation dataset, dubbed SAMRS. SAMRS totally possesses 105,090 images and 1,668,241 instances, surpassing existing high-resolution RS segmentation datasets in size by several orders of magnitude. It provides object category, location, and instance information that can be used for semantic segmentation, instance segmentation, and object detection, either individually or in combination. We also provide a comprehensive analysis of SAMRS from various aspects. Moreover, preliminary experiments highlight the importance of conducting segmentation pre-training with SAMRS to address task discrepancies and alleviate the limitations posed by limited training data during fine-tuning. The code and dataset will be available at https://github.com/ViTAE-Transformer/SAMRS. |
Accep...Accepted by NeurIPS 2023 Datasets and Benchmarks Track |
| Efficient Remote Sensing Segmentation With Generative Adversarial Transformer | 2023-10-02 | ShowMost deep learning methods that achieve high segmentation accuracy require deep network architectures that are too heavy and complex to run on embedded devices with limited storage and memory space. To address this issue, this paper proposes an efficient Generative Adversarial Transfomer (GATrans) for achieving high-precision semantic segmentation while maintaining an extremely efficient size. The framework utilizes a Global Transformer Network (GTNet) as the generator, efficiently extracting multi-level features through residual connections. GTNet employs global transformer blocks with progressively linear computational complexity to reassign global features based on a learnable similarity function. To focus on object-level and pixel-level information, the GATrans optimizes the objective function by combining structural similarity losses. We validate the effectiveness of our approach through extensive experiments on the Vaihingen dataset, achieving an average F1 score of 90.17% and an overall accuracy of 91.92%. |
|
| Semantic Labeling of High Resolution Images Using EfficientUNets and Transformers | 2022-06-22 | ShowSemantic segmentation necessitates approaches that learn high-level characteristics while dealing with enormous amounts of data. Convolutional neural networks (CNNs) can learn unique and adaptive features to achieve this aim. However, due to the large size and high spatial resolution of remote sensing images, these networks cannot analyze an entire scene efficiently. Recently, deep transformers have proven their capability to record global interactions between different objects in the image. In this paper, we propose a new segmentation model that combines convolutional neural networks with transformers, and show that this mixture of local and global feature extraction techniques provides significant advantages in remote sensing segmentation. In addition, the proposed model includes two fusion layers that are designed to represent multi-modal inputs and output of the network efficiently. The input fusion layer extracts feature maps summarizing the relationship between image content and elevation maps (DSM). The output fusion layer uses a novel multi-task segmentation strategy where class labels are identified using class-specific feature extraction layers and loss functions. Finally, a fast-marching method is used to convert all unidentified class labels to their closest known neighbors. Our results demonstrate that the proposed methodology improves segmentation accuracy compared to state-of-the-art techniques. |
|
| Evaluating Self and Semi-Supervised Methods for Remote Sensing Segmentation Tasks | 2022-06-19 | ShowSelf- and semi-supervised machine learning techniques leverage unlabeled data for improving downstream task performance. These methods are especially valuable for remote sensing tasks where producing labeled ground truth datasets can be prohibitively expensive but there is easy access to a wealth of unlabeled imagery. We perform a rigorous evaluation of SimCLR, a self-supervised method, and FixMatch, a semi-supervised method, on three remote sensing tasks: riverbed segmentation, land cover mapping, and flood mapping. We quantify performance improvements on these remote sensing segmentation tasks when additional imagery outside of the original supervised dataset is made available for training. We also design experiments to test the effectiveness of these techniques when the test set is domain shifted to sample different geographic areas compared to the training and validation sets. We find that such techniques significantly improve generalization performance when labeled data is limited and there are geographic domain shifts between the training data and the validation/test data. |
|
| How Useful is Region-based Classification of Remote Sensing Images in a Deep Learning Framework? | 2016-09-22 | ShowIn this paper, we investigate the impact of segmentation algorithms as a preprocessing step for classification of remote sensing images in a deep learning framework. Especially, we address the issue of segmenting the image into regions to be classified using pre-trained deep neural networks as feature extractors for an SVM-based classifier. An efficient segmentation as a preprocessing step helps learning by adding a spatially-coherent structure to the data. Therefore, we compare algorithms producing superpixels with more traditional remote sensing segmentation algorithms and measure the variation in terms of classification accuracy. We establish that superpixel algorithms allow for a better classification accuracy as a homogenous and compact segmentation favors better generalization of the training samples. |
IEEE ...IEEE International Geosciences and Remote Sensing Symposium (IGARSS), Jul 2016, Beijing, China |
| Title | Date | Abstract | Comment |
|---|---|---|---|
| RS-HyRe-R1: A Hybrid Reward Mechanism to Overcome Perceptual Inertia for Remote Sensing Images Understanding | 2026-04-19 | ShowReinforcement learning (RL) post-training substantially improves remote sensing vision-language models (RS-VLMs). However, when handling complex remote sensing imagery (RSI) requiring exhaustive visual scanning, models tend to rely on localized salient cues for rapid inference. We term this RL-induced bias "perceptual inertia". Driven by reward maximization, models favor quick outcome fitting, leading to two limitations: cognitively, overreliance on specific features impedes complete evidence construction; operationally, models struggle to flexibly shift visual focus across tasks. To address this bias and encourage comprehensive visual evidence mining, we propose RS-HyRe-R1, a hybrid reward framework for RSI understanding. It introduces: (1) a spatial reasoning activation reward that enforces structured visual reasoning; (2) a perception correctness reward that provides adaptive quality anchors across RS tasks, ensuring accurate geometric and semantic alignment; and (3) a visual-semantic path evolution reward that penalizes repetitive reasoning and promotes exploration of complementary cues to build richer evidence chains. Experiments show RS-HyRe-R1 effectively mitigates "perceptual inertia", encouraging deeper, more diverse reasoning. With only 3B parameters, it achieves state-of-the-art performance on REC, OVD, and VQA tasks, outperforming models up to 7B parameters. It also demonstrates strong zero-shot generalization, surpassing the second-best model by 3.16%, 3.97%, and 2.72% on VQA, OVD, and REC, respectively. Code and datasets are available at https://github.com/geox-lab/RS-HyRe-R1. |
|
| Continual Vision-Language Learning for Remote Sensing: Benchmarking and Analysis | 2026-04-01 | ShowCurrent remote sensing vision-language models (RS VLMs) demonstrate impressive performance in image interpretation but rely on static training data, limiting their ability to accommodate continuously emerging sensing modalities and downstream tasks. This exposes a fundamental challenge: enabling RS VLMs to continually adapt without catastrophic forgetting. Despite its practical importance, the continual learning capability of RS VLMs remains underexplored, and no dedicated benchmark currently exists. In this work, we present CLeaRS, a comprehensive benchmark for continual vision-language learning in remote sensing. CLeaRS comprises 10 curated subsets with over 207k image-text pairs, spanning diverse interpretation tasks, sensing modalities, and application scenarios. We further define three evaluation protocols: long-horizon, modality-incremental, and task-incremental settings, to systematically assess continual adaptation. Extensive benchmarking of diverse vision-language models reveals catastrophic forgetting across all settings. Moreover, representative continual learning methods, when adapted to RS VLMs, exhibit limited effectiveness in handling task, instruction, and modality transitions. Our findings underscore the need for developing continual learning methods tailored to RS VLMs. |
23 pa...23 pages, 7 figures, 9 tables |
| OmniEarth: A Benchmark for Evaluating Vision-Language Models in Geospatial Tasks | 2026-03-10 | ShowVision-Language Models (VLMs) have demonstrated effective perception and reasoning capabilities on general-domain tasks, leading to growing interest in their application to Earth observation. However, a systematic benchmark for comprehensively evaluating remote sensing vision-language models (RSVLMs) remains lacking. To address this gap, we introduce OmniEarth, a benchmark for evaluating RSVLMs under realistic Earth observation scenarios. OmniEarth organizes tasks along three capability dimensions: perception, reasoning, and robustness. It defines 28 fine-grained tasks covering multi-source sensing data and diverse geospatial contexts. The benchmark supports two task formulations: multiple-choice VQA and open-ended VQA. The latter includes pure text outputs for captioning tasks, bounding box outputs for visual grounding tasks, and mask outputs for segmentation tasks. To reduce linguistic bias and examine whether model predictions rely on visual evidence, OmniEarth adopts a blind test protocol and a quintuple semantic consistency requirement. OmniEarth includes 9,275 carefully quality-controlled images, including proprietary satellite imagery from Jilin-1 (JL-1), along with 44,210 manually verified instructions. We conduct a systematic evaluation of contrastive learning-based models, general closed-source and open-source VLMs, as well as RSVLMs. Results show that existing VLMs still struggle with geospatially complex tasks, revealing clear gaps that need to be addressed for remote sensing applications. OmniEarth is publicly available at https://huggingface.co/datasets/sjeeudd/OmniEarth. |
|
| GeoReason: Aligning Thinking And Answering In Remote Sensing Vision-Language Models Via Logical Consistency Reinforcement Learning | 2026-01-08 | ShowThe evolution of Remote Sensing Vision-Language Models(RS-VLMs) emphasizes the importance of transitioning from perception-centric recognition toward high-level deductive reasoning to enhance cognitive reliability in complex spatial tasks. However, current models often suffer from logical hallucinations, where correct answers are derived from flawed reasoning chains or rely on positional shortcuts rather than spatial logic. This decoupling undermines reliability in strategic spatial decision-making. To address this, we present GeoReason, a framework designed to synchronize internal thinking with final decisions. We first construct GeoReason-Bench, a logic-driven dataset containing 4,000 reasoning trajectories synthesized from geometric primitives and expert knowledge. We then formulate a two-stage training strategy: (1) Supervised Knowledge Initialization to equip the model with reasoning syntax and domain expertise, and (2) Consistency-Aware Reinforcement Learning to refine deductive reliability. This second stage integrates a novel Logical Consistency Reward, which penalizes logical drift via an option permutation strategy to anchor decisions in verifiable reasoning traces. Experimental results demonstrate that our framework significantly enhances the cognitive reliability and interpretability of RS-VLMs, achieving state-of-the-art performance compared to other advanced methods. |
|
| AirSpatialBot: A Spatially-Aware Aerial Agent for Fine-Grained Vehicle Attribute Recognization and Retrieval | 2026-01-04 | ShowDespite notable advancements in remote sensing vision-language models (VLMs), existing models often struggle with spatial understanding, limiting their effectiveness in real-world applications. To push the boundaries of VLMs in remote sensing, we specifically address vehicle imagery captured by drones and introduce a spatially-aware dataset AirSpatial, which comprises over 206K instructions and introduces two novel tasks: Spatial Grounding and Spatial Question Answering. It is also the first remote sensing grounding dataset to provide 3DBB. To effectively leverage existing image understanding of VLMs to spatial domains, we adopt a two-stage training strategy comprising Image Understanding Pre-training and Spatial Understanding Fine-tuning. Utilizing this trained spatially-aware VLM, we develop an aerial agent, AirSpatialBot, which is capable of fine-grained vehicle attribute recognition and retrieval. By dynamically integrating task planning, image understanding, spatial understanding, and task execution capabilities, AirSpatialBot adapts to diverse query requirements. Experimental results validate the effectiveness of our approach, revealing the spatial limitations of existing VLMs while providing valuable insights. The model, code, and datasets will be released at https://github.com/VisionXLab/AirSpatialBot |
12 pages, 9 figures |
| FUSE-RSVLM: Feature Fusion Vision-Language Model for Remote Sensing | 2025-12-30 | ShowLarge vision-language models (VLMs) exhibit strong performance across various tasks. However, these VLMs encounter significant challenges when applied to the remote sensing domain due to the inherent differences between remote sensing images and natural images. Existing remote sensing VLMs often fail to extract fine-grained visual features and suffer from visual forgetting during deep language processing. To address this, we introduce MF-RSVLM, a Multi-Feature Fusion Remote Sensing Vision--Language Model that effectively extracts and fuses visual features for RS understanding. MF-RSVLM learns multi-scale visual representations and combines global context with local details, improving the capture of small and complex structures in RS scenes. A recurrent visual feature injection scheme ensures the language model remains grounded in visual evidence and reduces visual forgetting during generation. Extensive experiments on diverse RS benchmarks show that MF-RSVLM achieves state-of-the-art or highly competitive performance across remote sensing classification, image captioning, and VQA tasks. Our code is publicly available at https://github.com/Yunkaidang/RSVLM. |
|
| ISS-Geo142: A Benchmark for Geolocating Astronaut Photography from the International Space Station | 2025-11-20 | ShowThis paper introduces ISS-Geo142, a curated benchmark for geolocating astronaut photography captured from the International Space Station (ISS). Although the ISS position at capture time is known precisely, the specific Earth locations depicted in these images are typically not directly georeferenced, making automated localization non-trivial. ISS-Geo142 consists of 142 images with associated metadata and manually determined geographic locations, spanning a range of spatial scales and scene types. On top of this benchmark, we implement and evaluate three geolocation pipelines: a neural network based approach (NN-Geo) using VGG16 features and cross-correlation over map-derived Areas of Interest (AOIs), a Scale-Invariant Feature Transform based pipeline (SIFT-Match) using sliding-window feature matching on stitched high-resolution AOIs, and TerraByte, an AI system built around a GPT-4 model with vision capabilities that jointly reasons over image content and ISS coordinates. On ISS-Geo142, NN-Geo achieves a match for 75.52% of the images under our evaluation protocol, SIFT-Match attains high precision on structurally rich scenes at substantial computational cost, and TerraByte establishes the strongest overall baseline, correctly geolocating approximately 90% of the images while also producing human-readable geographic descriptions. The methods and experiments were originally developed in 2023; this manuscript is a revised and extended version that situates the work relative to subsequent advances in cross-view geo-localization and remote-sensing vision--language models. Taken together, ISS-Geo142 and these three pipelines provide a concrete, historically grounded benchmark for future work on ISS image geolocation. |
|
| Few-Shot Adaptation Benchmark for Remote Sensing Vision-Language Models | 2025-10-08 | ShowRemote Sensing Vision-Language Models (RSVLMs) have shown remarkable potential thanks to large-scale pretraining, achieving strong zero-shot performance on various tasks. However, their ability to generalize in low-data regimes, such as few-shot learning, remains insufficiently explored. In this work, we present the first structured benchmark for evaluating few-shot adaptation methods on RSVLMs. We conduct comprehensive experiments across ten remote sensing scene classification datasets, applying five widely used few-shot adaptation strategies to three state-of-the-art RSVLMs with varying backbones. Our findings reveal that models with similar zero-shot performance can exhibit markedly different behavior under few-shot adaptation, with some RSVLMs being inherently more amenable to such adaptation than others. The variability of performance and the absence of a clear winner among existing methods highlight the need for the development of more robust methods for few-shot adaptation tailored to RS. To facilitate future research, we provide a reproducible benchmarking framework and open-source code to systematically evaluate RSVLMs under few-shot conditions. The source code is publicly available on Github: https://github.com/elkhouryk/fewshot_RSVLMs |
|
| DescribeEarth: Describe Anything for Remote Sensing Images | 2025-09-30 | ShowAutomated textual description of remote sensing images is crucial for unlocking their full potential in diverse applications, from environmental monitoring to urban planning and disaster management. However, existing studies in remote sensing image captioning primarily focus on the image level, lacking object-level fine-grained interpretation, which prevents the full utilization and transformation of the rich semantic and structural information contained in remote sensing images. To address this limitation, we propose Geo-DLC, a novel task of object-level fine-grained image captioning for remote sensing. To support this task, we construct DE-Dataset, a large-scale dataset contains 25 categories and 261,806 annotated instances with detailed descriptions of object attributes, relationships, and contexts. Furthermore, we introduce DE-Benchmark, a LLM-assisted question-answering based evaluation suite designed to systematically measure model capabilities on the Geo-DLC task. We also present DescribeEarth, a Multi-modal Large Language Model (MLLM) architecture explicitly designed for Geo-DLC, which integrates a scale-adaptive focal strategy and a domain-guided fusion module leveraging remote sensing vision-language model features to encode high-resolution details and remote sensing category priors while maintaining global context. Our DescribeEarth model consistently outperforms state-of-the-art general MLLMs on DE-Benchmark, demonstrating superior factual accuracy, descriptive richness, and grammatical soundness, particularly in capturing intrinsic object features and surrounding environmental attributes across simple, complex, and even out-of-distribution remote sensing scenarios. All data, code and weights are released at https://github.com/earth-insights/DescribeEarth. |
|
| RSVLM-QA: A Benchmark Dataset for Remote Sensing Vision Language Model-based Question Answering | 2025-08-11 | ShowVisual Question Answering (VQA) in remote sensing (RS) is pivotal for interpreting Earth observation data. However, existing RS VQA datasets are constrained by limitations in annotation richness, question diversity, and the assessment of specific reasoning capabilities. This paper introduces RSVLM-QA dataset, a new large-scale, content-rich VQA dataset for the RS domain. RSVLM-QA is constructed by integrating data from several prominent RS segmentation and detection datasets: WHU, LoveDA, INRIA, and iSAID. We employ an innovative dual-track annotation generation pipeline. Firstly, we leverage Large Language Models (LLMs), specifically GPT-4.1, with meticulously designed prompts to automatically generate a suite of detailed annotations including image captions, spatial relations, and semantic tags, alongside complex caption-based VQA pairs. Secondly, to address the challenging task of object counting in RS imagery, we have developed a specialized automated process that extracts object counts directly from the original segmentation data; GPT-4.1 then formulates natural language answers from these counts, which are paired with preset question templates to create counting QA pairs. RSVLM-QA comprises 13,820 images and 162,373 VQA pairs, featuring extensive annotations and diverse question types. We provide a detailed statistical analysis of the dataset and a comparison with existing RS VQA benchmarks, highlighting the superior depth and breadth of RSVLM-QA's annotations. Furthermore, we conduct benchmark experiments on Six mainstream Vision Language Models (VLMs), demonstrating that RSVLM-QA effectively evaluates and challenges the understanding and reasoning abilities of current VLMs in the RS domain. We believe RSVLM-QA will serve as a pivotal resource for the RS VQA and VLM research communities, poised to catalyze advancements in the field. |
This ...This paper has been accepted to the proceedings of the 33rd ACM International Multimedia Conference (ACM Multimedia 2025) |
| DeltaVLM: Interactive Remote Sensing Image Change Analysis via Instruction-guided Difference Perception | 2025-07-30 | ShowAccurate interpretation of land-cover changes in multi-temporal satellite imagery is critical for real-world scenarios. However, existing methods typically provide only one-shot change masks or static captions, limiting their ability to support interactive, query-driven analysis. In this work, we introduce remote sensing image change analysis (RSICA) as a new paradigm that combines the strengths of change detection and visual question answering to enable multi-turn, instruction-guided exploration of changes in bi-temporal remote sensing images. To support this task, we construct ChangeChat-105k, a large-scale instruction-following dataset, generated through a hybrid rule-based and GPT-assisted process, covering six interaction types: change captioning, classification, quantification, localization, open-ended question answering, and multi-turn dialogues. Building on this dataset, we propose DeltaVLM, an end-to-end architecture tailored for interactive RSICA. DeltaVLM features three innovations: (1) a fine-tuned bi-temporal vision encoder to capture temporal differences; (2) a visual difference perception module with a cross-semantic relation measuring (CSRM) mechanism to interpret changes; and (3) an instruction-guided Q-former to effectively extract query-relevant difference information from visual changes, aligning them with textual instructions. We train DeltaVLM on ChangeChat-105k using a frozen large language model, adapting only the vision and alignment modules to optimize efficiency. Extensive experiments and ablation studies demonstrate that DeltaVLM achieves state-of-the-art performance on both single-turn captioning and multi-turn interactive change analysis, outperforming existing multimodal large language models and remote sensing vision-language models. Code, dataset and pre-trained weights are available at https://github.com/hanlinwu/DeltaVLM. |
12 pa...12 pages, 5 figures. Submitted to IEEE Transactions on Geoscience and Remote Sensing (TGRS). Code and dataset are available at https://github.com/hanlinwu/DeltaVLM |
| Enhancing Remote Sensing Vision-Language Models Through MLLM and LLM-Based High-Quality Image-Text Dataset Generation | 2025-07-22 | ShowThe application of Vision-language foundation models (VLFMs) to remote sensing (RS) imagery has garnered significant attention due to their superior capability in various downstream tasks. A key challenge lies in the scarcity of high-quality, large-scale, image-text paired training data. Recently, several works introduced extensive image-text datasets for RS and trained their VLFMs. However, due to the rudimentary methods used for generating captions, the quality of datasets is suboptimal, requiring larger volumes of training data, while only yielding modest performance improvements. In this paper, we propose a two-stage method named MpGI(Multi-Perspective Generation and Integration) for generating high-quality text captions for RS images. Firstly, we generate distinct and detailed descriptions from different perspectives using Rule-MLLM(Multimodal Large Language Model) Relay Generation and MLLMs generation methods. Next, we utilize Large Language Models (LLMs) to integrate these diverse descriptions into comprehensive captions, capturing details from multiple perspectives. Finally, we have created the HQRS-IT-210K dataset, including about 210,000 RS images and 1.3 million captions. We fine-tuned two VLFMs using our dataset: CLIP, a discriminative model, and CoCa, an image-to-text generative model. This process resulted in our proposed HQRS-CLIP and RS-CoCa models. Experimental results demonstrate that HQRS-CLIP surpassed the previous SOTA RS CLIP model in various downstream tasks while using only 4.2% of the training data. RS-CoCa outperforms other advanced approaches across benchmark datasets and can generate captions for RS images that rival or even exceed manual annotations. Dataset, pre-trained models, and codes will be released at https://github.com/YiguoHe/HQRS-210K-and-HQRS-CLIP. |
SUBMI...SUBMIT TO IEEE TRANSACTIONS |
| Mind the Modality Gap: Towards a Remote Sensing Vision-Language Model via Cross-modal Alignment | 2025-07-18 | ShowDeep Learning (DL) is undergoing a paradigm shift with the emergence of foundation models. In this work, we focus on Contrastive Language-Image Pre-training (CLIP), a Vision-Language foundation model that achieves high accuracy across various image classification tasks and often rivals fully supervised baselines, despite not being explicitly trained for those tasks. Nevertheless, there are still domains where zero-shot CLIP performance is far from optimal, such as Remote Sensing (RS) and medical imagery. These domains do not only exhibit fundamentally different distributions compared to natural images, but also commonly rely on complementary modalities, beyond RGB, to derive meaningful insights. To this end, we propose a methodology to align distinct RS image modalities with the visual and textual modalities of CLIP. Our two-stage procedure addresses the aforementioned distribution shift, extends the zero-shot capabilities of CLIP and enriches CLIP's shared embedding space with domain-specific knowledge. Initially, we robustly fine-tune CLIP according to the PAINT (Ilharco et al., 2022) patching protocol, in order to deal with the distribution shift. Building upon this foundation, we facilitate the cross-modal alignment of a RS modality encoder by distilling knowledge from the CLIP visual and textual encoders. We empirically show that both patching and cross-modal alignment translate to significant performance gains, across several RS imagery classification and cross-modal retrieval benchmark datasets. Notably, these enhancements are achieved without the reliance on textual descriptions, without introducing any task-specific parameters, without training from scratch and without catastrophic forgetting. We make our code implementation and weights for all experiments publicly available at https://github.com/Orion-AI-Lab/MindTheModalityGap. |
Accep...Accepted at the ISPRS Journal of Photogrammetry and Remote Sensing. Our code implementation and weights for all experiments are publicly available at https://github.com/Orion-AI-Lab/MindTheModalityGap |
| A Survey on Remote Sensing Foundation Models: From Vision to Multimodality | 2025-03-28 | ShowThe rapid advancement of remote sensing foundation models, particularly vision and multimodal models, has significantly enhanced the capabilities of intelligent geospatial data interpretation. These models combine various data modalities, such as optical, radar, and LiDAR imagery, with textual and geographic information, enabling more comprehensive analysis and understanding of remote sensing data. The integration of multiple modalities allows for improved performance in tasks like object detection, land cover classification, and change detection, which are often challenged by the complex and heterogeneous nature of remote sensing data. However, despite these advancements, several challenges remain. The diversity in data types, the need for large-scale annotated datasets, and the complexity of multimodal fusion techniques pose significant obstacles to the effective deployment of these models. Moreover, the computational demands of training and fine-tuning multimodal models require significant resources, further complicating their practical application in remote sensing image interpretation tasks. This paper provides a comprehensive review of the state-of-the-art in vision and multimodal foundation models for remote sensing, focusing on their architecture, training methods, datasets and application scenarios. We discuss the key challenges these models face, such as data alignment, cross-modal transfer learning, and scalability, while also identifying emerging research directions aimed at overcoming these limitations. Our goal is to provide a clear understanding of the current landscape of remote sensing foundation models and inspire future research that can push the boundaries of what these models can achieve in real-world applications. The list of resources collected by the paper can be found in the https://github.com/IRIP-BUAA/A-Review-for-remote-sensing-vision-language-models. |
|
| Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification | 2025-01-07 | ShowVision-Language Models for remote sensing have shown promising uses thanks to their extensive pretraining. However, their conventional usage in zero-shot scene classification methods still involves dividing large images into patches and making independent predictions, i.e., inductive inference, thereby limiting their effectiveness by ignoring valuable contextual information. Our approach tackles this issue by utilizing initial predictions based on text prompting and patch affinity relationships from the image encoder to enhance zero-shot capabilities through transductive inference, all without the need for supervision and at a minor computational cost. Experiments on 10 remote sensing datasets with state-of-the-art Vision-Language Models demonstrate significant accuracy improvements over inductive zero-shot classification. Our source code is publicly available on Github: https://github.com/elkhouryk/RS-TransCLIP |
Accep...Accepted at ICASSP 2025 |
| RSUniVLM: A Unified Vision Language Model for Remote Sensing via Granularity-oriented Mixture of Experts | 2024-12-10 | ShowRemote Sensing Vision-Language Models (RS VLMs) have made much progress in the tasks of remote sensing (RS) image comprehension. While performing well in multi-modal reasoning and multi-turn conversations, the existing models lack pixel-level understanding and struggle with multi-image inputs. In this work, we propose RSUniVLM, a unified, end-to-end RS VLM designed for comprehensive vision understanding across multiple granularity, including image-level, region-level, and pixel-level tasks. RSUniVLM also performs effectively in multi-image analysis, with instances of change detection and change captioning. To enhance the model's ability to capture visual information at different levels without increasing model size, we design a novel architecture called Granularity-oriented Mixture of Experts to constraint the model to about 1 billion parameters. We also construct a large-scale RS instruction-following dataset based on a variety of existing datasets in both RS and general domain, encompassing various tasks such as object localization, visual question answering, and semantic segmentation. Substantial experiments have been conducted to validate the superiority of the proposed RSUniVLM up to state-of-the-art across various RS tasks. Code and model will be available at \href{https://github.com/xuliu-cyber/RSUniVLM}{here}. |
|
| Aquila: A Hierarchically Aligned Visual-Language Model for Enhanced Remote Sensing Image Comprehension | 2024-11-09 | ShowRecently, large vision language models (VLMs) have made significant strides in visual language capabilities through visual instruction tuning, showing great promise in the field of remote sensing image interpretation. However, existing remote sensing vision language models (RSVLMs) often fall short in capturing the complex characteristics of remote sensing scenes, as they typically rely on low resolution, single scale visual features and simplistic methods to map visual features to language features. In this paper, we present Aquila, an advanced visual language foundation model designed to enable richer visual feature representation and more precise visual-language feature alignment for remote sensing images. Our approach introduces a learnable Hierarchical Spatial Feature Integration (SFI) module that supports high resolution image inputs and aggregates multi scale visual features, allowing for the detailed representation of complex visual information. Additionally, the SFI module is repeatedly integrated into the layers of the large language model (LLM) to achieve deep visual language feature alignment, without compromising the model's performance in natural language processing tasks. These innovations, capturing detailed visual effects through higher resolution and multi scale input, and enhancing feature alignment significantly improve the model's ability to learn from image text data. We validate the effectiveness of Aquila through extensive quantitative experiments and qualitative analyses, demonstrating its superior performance. |
|
| RSGPT: A Remote Sensing Vision Language Model and Benchmark | 2023-07-28 | ShowThe emergence of large-scale large language models, with GPT-4 as a prominent example, has significantly propelled the rapid advancement of artificial general intelligence and sparked the revolution of Artificial Intelligence 2.0. In the realm of remote sensing (RS), there is a growing interest in developing large vision language models (VLMs) specifically tailored for data analysis in this domain. However, current research predominantly revolves around visual recognition tasks, lacking comprehensive, large-scale image-text datasets that are aligned and suitable for training large VLMs, which poses significant challenges to effectively training such models for RS applications. In computer vision, recent research has demonstrated that fine-tuning large vision language models on small-scale, high-quality datasets can yield impressive performance in visual and language understanding. These results are comparable to state-of-the-art VLMs trained from scratch on massive amounts of data, such as GPT-4. Inspired by this captivating idea, in this work, we build a high-quality Remote Sensing Image Captioning dataset (RSICap) that facilitates the development of large VLMs in the RS field. Unlike previous RS datasets that either employ model-generated captions or short descriptions, RSICap comprises 2,585 human-annotated captions with rich and high-quality information. This dataset offers detailed descriptions for each image, encompassing scene descriptions (e.g., residential area, airport, or farmland) as well as object information (e.g., color, shape, quantity, absolute position, etc). To facilitate the evaluation of VLMs in the field of RS, we also provide a benchmark evaluation dataset called RSIEval. This dataset consists of human-annotated captions and visual question-answer pairs, allowing for a comprehensive assessment of VLMs in the context of RS. |
| Title | Date | Abstract | Comment |
|---|---|---|---|
| Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs | 2026-05-01 | ShowWhile autoregressive Large Vision-Language Models (LVLMs) demonstrate remarkable proficiency in multimodal tasks, they face a "Visual Signal Dilution" phenomenon, where the accumulation of textual history expands the attention partition function, causing visual attention to decay inversely with generated sequence length. To counteract this, we propose Persistent Visual Memory (PVM), a lightweight learnable module designed to ensure sustained, on-demand visual perception. Integrated as a parallel branch alongside the Feed-Forward Network (FFN) in LVLMs, PVM establishes a distance-agnostic retrieval pathway that directly provides visual embeddings for precise visual perception, thereby structurally mitigating the signal suppression inherent to deep generation. Extensive experiments on Qwen3-VL models demonstrate that PVM brings notable improvements with negligible parameter overhead, delivering consistent average accuracy gains across both 4B and 8B scales, particularly in complex reasoning tasks that demand persistent visual perception. Furthermore, in-depth analysis reveals that PVM can resist length-induced signal decay and accelerate internal prediction convergence. |
|
| ScreenParse: Moving Beyond Sparse Grounding with Complete Screen Parsing Supervision | 2026-05-01 | ShowModern computer-use agents (CUA) must perceive a screen as a structured state, what elements are visible, where they are, and what text they contain, before they can reliably ground instructions and act. Yet, most available grounding datasets provide sparse supervision, with insufficient and low-diversity labels that annotate only a small subset of task-relevant elements per screen, which limits both coverage and generalization; moreover, practical deployment requires efficiency to enable low-latency, on-device use. We introduce ScreenParse, a large-scale dataset for complete screen parsing, with dense annotations of all visible UI elements (boxes, 55-class types, and text) across 771K web screenshots (21M elements). ScreenParse is generated by Webshot, an automated, scalable pipeline that renders diverse urls, extracts annotations and applies VLM-based relabeling and quality filtering. Using ScreenParse, we train ScreenVLM, a compact, 316M-parameter vision language model (VLM) that decodes a compact ScreenTag markup representation with a structure-aware loss that upweights structure-critical tokens. ScreenVLM substantially outperforms much larger foundation VLMs on dense parsing (e.g., 0.592 vs. 0.294 PageIoU on ScreenParse) and shows strong transfer to public benchmarks. Moreover, finetuning foundation VLMs on ScreenParse consistently improves their grounding performance, suggesting that dense screen supervision provides transferable structural priors for UI understanding. Project page: https://saidgurbuz.github.io/screenparse/. |
Accep...Accepted at ICML 2026. 28 pages, 15 figures |
| Make Your LVLM KV Cache More Lightweight | 2026-05-01 | ShowKey-Value (KV) cache has become a de facto component of modern Large Vision-Language Models (LVLMs) for inference. While it enhances decoding efficiency in Large Language Models (LLMs), its direct adoption in LVLMs introduces substantial GPU memory overhead due to the large number of vision tokens processed during the prefill stage. To tackle this problem, we propose LightKV, a novel approach that reduces KV cache size by exploiting the redundancy among vision-token embeddings. Guided by text prompts, LightKV employs cross-modality message passing to aggregate informative messages across vision tokens and progressively compress them during prefill. This prompt-aware guidance distinguishes our method from prior vision-only compression strategies. We evaluate LightKV on eight open-source LVLMs across eight public benchmark datasets, e.g., MME and SeedBench. Experimental results demonstrate that with only 55% of the original vision tokens, LightKV (a) halves the vision-token KV cache size, (b) reduces computation by up to 40%, and (c) preserves general-purpose performance while significantly outperforming existing baselines. |
Accep...Accepted to Transactions on Machine Learning Research (TMLR), 2026 |
| STARE: Step-wise Temporal Alignment and Red-teaming Engine for Multi-modal Toxicity Attack | 2026-05-01 | ShowRed-teaming Vision-Language Models is essential for identifying vulnerabilities where adversarial image-text inputs trigger toxic outputs. Existing approaches treat image generation as a black box, returning only terminal toxicity scores and leaving open the question of when and how toxic semantics emerge during multi-step synthesis. We introduce STARE, a hierarchical reinforcement learning framework that treats the denoising trajectory itself as the attack surface, under a direct white-box T2I and query-only black-box VLM setting. By coupling a high-level prompt editor with low-level T2I fine-tuning via Group Relative Policy Optimization (GRPO), STARE attains a 68% improvement in Attack Success Rate over state-of-the-art black-box and white-box baselines. More importantly, this trajectory-level view surfaces the Optimization-Induced Phase Alignment phenomenon: vanilla models exhibit diffuse toxicity, whereas adversarial optimization concentrates conceptual harms into early semantic phases and detail-oriented harms into late refinement. Targeted perturbations of either window selectively suppress different toxicity categories, indicating that this temporal structure is a genuine causal handle rather than a side effect of the hierarchical design. The phenomenon turns toxicity formation from a chaotic process into a small set of predictable vulnerability windows, providing both a potent attack engine and a basis for phase-aware safety mechanisms. Content warning: This paper contains examples of toxic content that may be offensive or disturbing. |
|
| Adaptive Dual-Teacher Distillation with Subnetwork Rectification for Bridging Semantic Gaps in Black-Box Domain Adaptation | 2026-05-01 | ShowAssuming that neither source data nor source model parameters are accessible, black-box domain adaptation (BBDA) represents a highly practical yet challenging setting, where transferable knowledge is limited to the predictions of a black-box source model. Existing approaches exploit such knowledge via pseudo-label refinement or by leveraging vision-language models (ViLs), but they often fail to reconcile the inherent discrepancy between task-specific knowledge from black-box models and language-aligned semantic priors of ViLs, resulting in suboptimal integration and degraded adaptation performance. To address this challenge, we propose adaptive Dual-Teacher Distillation with Subnetwork Rectification (DDSR), a framework that explicitly reconciles these complementary yet inconsistent knowledge sources. DDSR employs an adaptive prediction fusion strategy to integrate predictions from the black-box source model and a ViL, generating reliable pseudo-labels for the target domain. A subnetwork-based regularization mechanism mitigates overfitting to noisy supervision by enforcing output consistency and gradient divergency. Furthermore, progressively improved target predictions iteratively refine both pseudo-labels and ViL prompts, enhancing semantic alignment. Finally, class-wise prototypes are used to further optimize target predictions via self-training. Extensive experiments on multiple benchmark datasets demonstrate that DDSR consistently outperforms state-of-the-art methods, including those with access to source data or source model parameters. |
Under Review |
| Intrinsic Gradient Suppression for Label-Noise Prompt Tuning in Vision-Language Models | 2026-05-01 | ShowContrastive vision-language models like CLIP exhibit remarkable zero-shot generalization. However, prompt tuning remains highly sensitive to label noise, as mislabeled samples generate disproportionately large gradients that can overwhelm pre-trained priors. We argue that because CLIP already provides a near-optimal initialization, adaptation should be inherently conservative, particularly against the extreme gradient updates common in noisy settings. To this end, we propose Double-Softmax Prompt Tuning (DSPT), a hyperparameter-free method for intrinsic gradient suppression. By applying a sequential probabilistic normalization, DSPT induces a self-adaptive saturation zone that suppresses gradients from high-error noisy samples while maintaining informative updates. We also provide both theoretical analysis and empirical evidence about how this mechanism achieves adaptive suppression. This design transforms ``gradient vanishing'', traditionally a training bottleneck, into a principled noise-filtering shield for label-noise prompt tuning. Extensive experiments confirm that this simple, drop-in design achieves state-of-the-art robustness across various noisy benchmarks, outperforming methods with complex architectures and handcrafted hyperparameters. |
|
| Jailbreaking Vision-Language Models Through the Visual Modality | 2026-05-01 | ShowThe visual modality of vision-language models (VLMs) is an underexplored attack surface for bypassing safety alignment. We introduce four jailbreak attacks exploiting the vision component: (1) encoding harmful instructions as visual symbol sequences with a decoding legend, (2) replacing harmful objects with benign substitutes (e.g., bomb -> banana) then prompting for harmful actions using the substitute term, (3) replacing harmful text in images (e.g., on book covers) with benign words while visual context preserves the original meaning, and (4) visual analogy puzzles whose solution requires inferring a prohibited concept. Evaluating across six frontier VLMs, our visual attacks bypass safety alignment and expose a cross-modality alignment gap: text-based safety training does not automatically generalize to harmful intent conveyed visually. For example, our visual cipher achieves 40.9% attack success on Claude-Haiku-4.5 versus 10.7% for an equivalent textual cipher. To further our insight into the attack mechanism, we present preliminary interpretability and mitigation results. These findings highlight that robust VLM alignment requires treating vision as a first-class target for safety post-training. |
Accep...Accepted to ICML 2026 |
| RoofNet: A Global Multimodal Dataset for Roof Material Classification | 2026-05-01 | ShowNatural disasters are increasing in frequency and severity, causing hundreds of billions of dollars in damage annually and posing growing threats to infrastructure and human livelihoods. Accurate data on roofing materials is critical for modeling building vulnerability to natural hazards such as earthquakes, floods, wildfires, and hurricanes, yet such data remain unavailable. To address this gap, we introduce RoofNet, the largest and most geographically diverse novel multimodal dataset to date, comprising over 51,500 samples from 184 geographically diverse sites pairing high-resolution Earth Observation (EO) imagery with curated text annotations for global roof material classification. RoofNet includes geographically diverse satellite imagery labeled with 14 key roofing types and is designed to enhance the fidelity of global exposure datasets through vision-language modeling (VLM). We sample EO tiles from climatically and architecturally distinct regions to construct a representative dataset. A subset of 6,000 images was annotated in collaboration with domain experts to fine-tune a VLM. We used geographic- and material-aware prompt tuning to enhance class separability. The fine-tuned model was then applied to the remaining EO tiles, with predictions refined through rule-based and human-in-the-loop verification. In addition to material labels, RoofNet provides rich metadata including roof shape, footprint area, solar panel presence, and indicators of mixed roofing materials (e.g., HVAC systems). The dataset used in earlier experiments has been removed due to licensing constraints related to imagery sources. Results based on this dataset should be interpreted with caution. Updated experiments using compliant data are in progress. |
v2: C...v2: Clarified data sourcing and removed references to imagery subject to licensing restrictions |
| Adapting Large VLMs with Iterative and Manual Instructions for Generative Low-light Enhancement | 2026-05-01 | ShowMost existing low-light image enhancement (LLIE) methods rely on pre-trained model priors, low-light inputs, or both, while neglecting the semantic guidance available from normal-light images. This limitation hinders their effectiveness in complex lighting conditions. In this paper, we propose VLM-IMI, a framework that adapts large vision-language models with iterative and manual instructions for generative LLIE. VLM-IMI mainly contains two branches: Normal-Light Instruction Prior Generation (NL-IPG) and Instruction-aware Light Enhancement Diffusion (IA-LED). The NL-IPG incorporates textual descriptions of the desired normal-light content as enhancement cues, enabling semantically informed restoration. IA-LED incorporates instruction priors from the NL-IPG to guide the diffusion process, enabling precise illumination enhancement. To effectively integrate cross-modal priors, we introduce a learnable instruction prior fusion module, which dynamically aligns and fuses image and text features, promoting the generation of detailed and semantically coherent outputs. During inference, as the ground-truth normal-light images are not available, we propose an inference with an iterative instructions strategy to refine textual instructions, progressively improving visual quality. Our VLM-IMI also inherently supports manual instruction control by allowing users to directly input custom instructions into the LLM to generate user-expected outputs. Experiments across diverse scenarios demonstrate that VLM-IMI outperforms SOTA methods in terms of perception and realism. The source code is available at: https://github.com/sunxiaoran01/VLM-IMI. |
11 pa...11 papers,8 figures, CVPR2026 Findings |
| Leveraging Vision-Language Models as Weak Annotators in Active Learning | 2026-05-01 | ShowActive learning aims to reduce annotation cost by selectively querying informative samples for supervision under a limited labeling budget. In this work, we investigate how vision-language models (VLMs) can be leveraged to further reduce the reliance on costly human annotation within the active learning paradigm. To this end, we find that the reliability of VLMs varies significantly with label granularity in fine-grained recognition tasks: they perform poorly on fine-grained labels but can provide accurate coarse-grained labels. Leveraging this property, we propose an active learning framework that combines fine-grained human annotations with coarse-grained VLM-generated weak labels through instance-wise label assignment. We further model the systematic noise in VLM-generated labels using a small set of trusted full labels. Experiments on CUB200 and FGVC-Aircraft show that the proposed framework consistently outperforms existing active learning methods under the same annotation budget. |
Accepted at ICIP2026 |
| Thinking in Text and Images: Interleaved Vision--Language Reasoning Traces for Long-Horizon Robot Manipulation | 2026-05-01 | ShowLong-horizon robotic manipulation requires plans that are both logically coherent and geometrically grounded. Existing Vision-Language-Action policies usually hide planning in latent states or expose only one modality: text-only chain-of-thought encodes causal order but misses spatial constraints, while visual prediction provides geometric cues but often remains local and semantically underconstrained. We introduce Interleaved Vision--Language Reasoning (IVLR), a policy framework built around \trace{}, an explicit intermediate representation that alternates textual subgoals with visual keyframes over the full task horizon. At test time, a single native multimodal transformer self-generates this global semantic-geometric trace from the initial observation and instruction, caches it, and conditions a closed-loop action decoder on the trace, original instruction, and current observation. Because standard robot datasets lack such traces, we construct pseudo-supervision by temporally segmenting demonstrations and captioning each stage with a vision-language model. Across simulated benchmarks for long-horizon manipulation and visual distribution shift, \method{} reaches 95.5% average success on LIBERO, including 92.4% on LIBERO-Long, and 59.4% overall success on SimplerEnv-WidowX. Ablations show that both modalities are necessary: without traces, LIBERO-Long success drops to 37.7%; text-only and vision-only traces reach 62.0% and 68.4%, while the full interleaved trace reaches 92.4%. Stress tests with execution perturbations and masked trace content show moderate degradation, suggesting that the trace can tolerate local corruption and moderate execution drift, but remains limited under stale or incorrect global plans. |
|
| InterChart: Benchmarking Visual Reasoning Across Decomposed and Distributed Chart Information | 2026-05-01 | ShowWe introduce InterChart, a diagnostic benchmark that evaluates how well vision-language models (VLMs) reason across multiple related charts, a task central to real-world applications such as scientific reporting, financial analysis, and public policy dashboards. Unlike prior benchmarks focusing on isolated, visually uniform charts, InterChart challenges models with diverse question types ranging from entity inference and trend correlation to numerical estimation and abstract multi-step reasoning grounded in 2-3 thematically or structurally related charts. We organize the benchmark into three tiers of increasing difficulty: (1) factual reasoning over individual charts, (2) integrative analysis across synthetically aligned chart sets, and (3) semantic inference over visually complex, real-world chart pairs. Our evaluation of state-of-the-art open- and closed-source VLMs reveals consistent and steep accuracy declines as chart complexity increases. We find that models perform better when we decompose multi-entity charts into simpler visual units, underscoring their struggles with cross-chart integration. By exposing these systematic limitations, InterChart provides a rigorous framework for advancing multimodal reasoning in complex, multi-visual environments. |
22 pa...22 pages, 8 figures, 14 tables. Accepted at IJCNLP-AACL 2025 |
| RTPrune: Reading-Twice Inspired Token Pruning for Efficient DeepSeek-OCR Inference | 2026-05-01 | ShowDeepSeek-OCR leverages visual-text compression to reduce long-text processing costs and accelerate inference, yet visual tokens remain prone to redundant textual and structural information. Moreover, current token pruning methods for conventional vision-language models (VLMs) fail to preserve textual fidelity due to improper compression mechanisms. By analyzing the decoding process of DeepSeek-OCR, we find that a distinct two-stage reading trajectory: the model initially prioritizes the majority of high-norm tokens, then subsequently redistributes its attention to the remaining ones. Motivated by this insight, we propose RTPrune, a two-stage token pruning method tailored for DeepSeek-OCR. In the first stage, we prioritize high-norm visual tokens that capture salient textual and structural information. In the second stage, the remaining tokens are paired and merged based on optimal transport theory to achieve efficient feature aggregation. We further introduce a dynamic pruning ratio that adapts to token similarity and textual density for OCR tasks, enabling a better efficiency-accuracy trade-off. Extensive experiments demonstrate state-of-the-art performance, as evidenced by 99.47% accuracy and 1.23$\times$ faster prefill on OmniDocBench, achieved with 84.25% token retention when applied to DeepSeek-OCR-Large. |
19p p...19p pages, accepted by ICML2026 |
| Odysseus: Scaling VLMs to 100+ Turn Decision-Making in Games via Reinforcement Learning | 2026-05-01 | ShowGiven the rapidly growing capabilities of vision-language models (VLMs), extending them to interactive decision-making tasks such as video games has emerged as a promising frontier. However, existing approaches either rely on large-scale supervised fine-tuning (SFT) on human trajectories or apply reinforcement learning (RL) only in relatively short-horizon settings (typically around 20--30 turns). In this work, we study RL-based training of VLMs for long-horizon decision-making in Super Mario Land, a visually grounded environment requiring 100+ turns of interaction with coordinated perception, reasoning, and action. We begin with a systematic investigation of key algorithmic components and propose an adapted variant of PPO with a lightweight turn-level critic, which substantially improves training stability and sample efficiency over critic-free methods such as GRPO and Reinforce++. We further show that pretrained VLMs provide strong action priors, significantly improving sample efficiency during RL training and reducing the need for manual design choices such as action engineering, compared to classical deep RL trained from scratch. Building on these insights, we introduce Odysseus, an open training framework for VLM agents, achieving substantial gains across multiple levels of the game and at least 3 times average game progresses than frontier models. Moreover, the trained models exhibit consistent improvements under both in-game and cross-game generalization settings, while maintaining general-domain capabilities. Overall, our results identify key ingredients for making RL stable and effective in long-horizon, multi-modal settings, and provide practical guidance for developing VLMs as embodied agents. |
|
| A Survey on Vision-Language-Action Models for Embodied AI | 2026-05-01 | ShowEmbodied AI is widely recognized as a cornerstone of artificial general intelligence (AGI) because it involves controlling embodied agents to perform tasks in the physical world. Building on the success of large language models (LLMs) and vision-language models (VLMs), a new category of multimodal models -- referred to as vision-language-action (VLA) models -- has emerged to address language-conditioned robotic tasks in embodied AI by leveraging their distinct ability to generate actions. The recent proliferation of VLAs necessitates a comprehensive survey to capture the rapidly evolving landscape. To this end, we present the first survey on VLAs for embodied AI. This work provides a detailed taxonomy of VLAs, organized into three major lines of research. The first line focuses on individual components of VLAs. The second line is dedicated to developing VLA-based control policies adept at predicting low-level actions. The third line comprises high-level task planners capable of decomposing long-horizon tasks into a sequence of subtasks, thereby guiding VLAs to follow more general user instructions. Furthermore, we provide an extensive summary of relevant resources, including datasets, simulators, and benchmarks. Finally, we discuss the challenges facing VLAs and outline promising future directions in embodied AI. A curated repository associated with this survey is available at: https://github.com/yueen-ma/Awesome-VLA. |
Proje...Project page: https://github.com/yueen-ma/Awesome-VLA |
| Prompt-Induced Score Variance in Zero-Shot Binary Vision-Language Safety Classification | 2026-05-01 | ShowSingle-prompt first-token probabilities from zero-shot vision-language model (VLM) safety classifiers are treated as decision scores, but we show they are unreliable under semantically equivalent prompt reformulation: even when the binary label is constrained to a fixed output position, equivalent prompts can induce materially different unsafe probabilities for the same sample. Across multimodal safety benchmarks and multiple VLM families, cross-prompt variance is strongly associated with prompt-level disagreement and higher error, making it a useful fragility diagnostic. A training-free mean ensemble improves NLL on all 14 dataset-model evaluation pairs and ECE on 12/14 relative to a train-selected single-prompt baseline, and wins more head-to-head NLL comparisons than labeled temperature scaling, Platt scaling, and isotonic regression applied to the same prompt. Ranking gains are consistent against the train-selected baseline on both AUROC and AUPRC, and against the full 15-prompt distribution remain consistent on AUPRC while softening on AUROC. Labeled calibration on top of the mean provides further gains when labels are available, identifying prompt averaging as a strong label-free first stage rather than a replacement for calibration. We frame this as a reliability stress test for zero-shot VLM first-token safety scores and recommend prompt-family evaluation with mean aggregation as a standard label-free reliability baseline. |
Prepr...Preprint. 19 pages, 5 figures |
| Online Self-Calibration Against Hallucination in Vision-Language Models | 2026-05-01 | ShowLarge Vision-Language Models (LVLMs) often suffer from hallucinations, generating descriptions that include visual details absent from the input image. Recent preference alignment methods typically rely on supervision distilled from stronger models such as GPT. However, this offline paradigm introduces a Supervision-Perception Mismatch: the student model is forced to align with fine-grained details beyond its perceptual capacity, learning to guess rather than to see. To obtain reliable self-supervision for online learning, we identify a Generative-Discriminative Gap within LVLMs, where models exhibit higher accuracy on discriminative verification than open-ended generation. Leveraging this capability, we propose \textbf{O}nline \textbf{S}elf-\textbf{CA}lib\textbf{R}ation (OSCAR), a framework that integrates Monte Carlo Tree Search with a Dual-Granularity Reward Mechanism to construct preference data and iteratively refines the model via Direct Preference Optimization. Extensive experiments demonstrate that OSCAR achieves state-of-the-art performance on hallucination benchmarks while improving general multimodal capabilities. |
IJCAI 2026 |
| Unlocking Zero-Shot Geospatial Reasoning via Indirect Rewards | 2026-04-30 | ShowTraining robust reasoning vision-language models (VLMs) in rare domains (such as geospatial) is fundamentally constrained by supervision scarcity. While raw geospatial imagery is abundant, the amount of task-direct supervision falls far behind that of common domains. In this work, we validate an important conclusion: indirect verifiable rewards, derived from seemingly unrelated metadata, are sufficient to induce sophisticated and generalizable geospatial reasoning across a wide range of downstream tasks (25+). We present Geo-R1 as one empirical instantiation of this paradigm. Rather than relying on limited task-specific annotations (i.e., direct rewards), Geo-R1 utilizes scalable, verifiable indirect proxy rewards based on cross-view alignment with metadata (geolocation information) to drive reinforcement learning at scale. Such indirect rewards successfully motivate the model to discover and internalize zero-shot geospatial reasoning across diverse tasks, achieving extraordinary zero-shot transfer on out-of-distribution benchmarks and even surpassing fully supervised specialists on certain benchmarks. These findings indicate that optimizing for indirect verifiable rewards may provide a scalable pathway to unlock generalized reasoning capabilities in rare domains with massive unlabeled data archives. Our code is availavle at: https://github.com/miniHuiHui/Geo-R1. |
ICML 2026 |
| CaTS-Bench: Can Language Models Describe Time Series? | 2026-04-30 | ShowTime series captioning, the task of describing time series in natural language, requires numeric and temporal reasoning, trend interpretation, and contextual understanding. Existing benchmarks, however, often rely on fully synthetic or generic captions, and typically neglect metadata and visual representations. We introduce CaTS-Bench, a comprehensive benchmark for Context-aware Time Series reasoning across 11 diverse domains, centered on a gold-standard evaluation set of 1746 human-rewritten captions that measure how effectively models translate numeric trends into immediately interpretable narratives. To address the scarcity of human-annotated data, we also propose a scalable pipeline for generating high-fidelity synthetic captions, the quality of which we validate. We evaluate leading Vision-Language Models on our benchmark, revealing that even proprietary models struggle to capture numeric nuances in temporal descriptions, while finetuning open-source models on synthetic data yields substantial performance gains. Finally, we release a diagnostic suite of 910 multiple-choice questions and use tailored numeric metrics to gauge time-series-specific reasoning capabilities, establishing CaTS-Bench as a reliable foundation for grounded, multimodal text generation in numeric domains. |
9 pag...9 pages, 6 figures, 4 tables in the main paper. Many more in the appendix |
| PhyCo: Learning Controllable Physical Priors for Generative Motion | 2026-04-30 | ShowModern video diffusion models excel at appearance synthesis but still struggle with physical consistency: objects drift, collisions lack realistic rebound, and material responses seldom match their underlying properties. We present PhyCo, a framework that introduces continuous, interpretable, and physically grounded control into video generation. Our approach integrates three key components: (i) a large-scale dataset of over 100K photorealistic simulation videos where friction, restitution, deformation, and force are systematically varied across diverse scenarios; (ii) physics-supervised fine-tuning of a pretrained diffusion model using a ControlNet conditioned on pixel-aligned physical property maps; and (iii) VLM-guided reward optimization, where a fine-tuned vision-language model evaluates generated videos with targeted physics queries and provides differentiable feedback. This combination enables a generative model to produce physically consistent and controllable outputs through variations in physical attributes-without any simulator or geometry reconstruction at inference. On the Physics-IQ benchmark, PhyCo significantly improves physical realism over strong baselines, and human studies confirm clearer and more faithful control over physical attributes. Our results demonstrate a scalable path toward physically consistent, controllable generative video models that generalize beyond synthetic training environments. |
CVPR ...CVPR 2026. Project Page: https://phyco-video.github.io/ |
| FreeOcc: Training-Free Embodied Open-Vocabulary Occupancy Prediction | 2026-04-30 | ShowExisting learning-based occupancy prediction methods rely on large-scale 3D annotations and generalize poorly across environments. We present FreeOcc, a training-free framework for open-vocabulary occupancy prediction from monocular or RGB-D sequences. Unlike prior approaches that require voxel-level supervision and ground-truth camera poses, FreeOcc operates without 3D annotations, pose ground truth, or any learning stage. FreeOcc incrementally builds a globally consistent occupancy map via a four-layer pipeline: a SLAM backbone estimates poses and sparse geometry; a geometrically consistent Gaussian update constructs dense 3D Gaussian maps; open-vocabulary semantics from off-the-shelf vision-language models are associated with Gaussian primitives; and a probabilistic Gaussian-to-occupancy projection produces dense voxel occupancy. Despite being entirely training-free and pose-agnostic, FreeOcc achieves over |
RSS 2026 |
| Benchmarking Large Vision-Language Models on Fine-Grained Image Tasks: A Comprehensive Evaluation | 2026-04-30 | ShowRecent advancements in Large Vision-Language Models (LVLMs) have demonstrated remarkable multimodal perception capabilities, garnering significant attention. While numerous evaluation studies have emerged, assessing LVLMs both holistically and on specialized tasks, fine-grained image tasks-fundamental to computer vision-remain largely unexplored. To fill this gap, we introduce a comprehensive fine-grained evaluation benchmark, i.e., FG-BMK, comprising 1.01 million questions and 0.33 million images. Our evaluation systematically examines LVLMs from both human-oriented and machine-oriented perspectives, focusing on their semantic recognition and fine-grained feature representation capabilities. Through extensive experiments on twelve representative LVLMs/VLMs, we uncover key findings regarding the influence of training paradigms, modality alignment, perturbation susceptibility, and fine-grained category reasoning on task performance. This work provides critical insights into the limitations of current LVLMs and offers guidance for future data construction and model design in the development of more advanced LVLMs. Our code is open-source and available at https://github.com/SEU-VIPGroup/FG-BMK. |
Accep...Accepted to ICLR 2026 |
| TransVLM: A Vision-Language Framework and Benchmark for Detecting Any Shot Transitions | 2026-04-30 | ShowTraditional Shot Boundary Detection (SBD) inherently struggles with complex transitions by formulating the task around isolated cut points, frequently yielding corrupted video shots. We address this fundamental limitation by formalizing the Shot Transition Detection (STD) task. Rather than searching for ambiguous points, STD explicitly detects the continuous temporal segments of transitions. To tackle this, we propose TransVLM, a Vision-Language Model (VLM) framework for STD. Unlike regular VLMs that predominantly rely on spatial semantics and struggle with fine-grained inter-shot dynamics, our method explicitly injects optical flow as a critical motion prior at the input stage. Through a simple yet effective feature-fusion strategy, TransVLM directly processes concatenated color and motion representations, significantly enhancing its temporal awareness without incurring any additional visual token overhead on the language backbone. To overcome the severe class imbalance in public data, we design a scalable data engine to synthesize diverse transition videos for robust training, alongside a comprehensive benchmark for STD. Extensive experiments demonstrate that TransVLM achieves superior overall performance, outperforming traditional heuristic methods, specialized spatiotemporal networks, and top-tier VLMs. This work has been deployed to production. For more related research, please visit HeyGen Research (https://www.heygen.com/research) and HeyGen Avatar-V (https://www.heygen.com/research/avatar-v-model). Project page: https://chence17.github.io/TransVLM/ |
This ...This work has been deployed to production. For more related research, please visit HeyGen Research (https://www.heygen.com/research) and HeyGen Avatar-V (https://www.heygen.com/research/avatar-v-model). Project page: https://chence17.github.io/TransVLM/ |
| FineState-Bench: Benchmarking State-Conditioned Grounding for Fine-grained GUI State Setting | 2026-04-30 | ShowDespite the rapid progress of large vision-language models (LVLMs), fine-grained, state-conditioned GUI interaction remains challenging. Current evaluations offer limited coverage, imprecise target-state definitions, and an overreliance on final-task success, obscuring where and why agents fail. To address this gap, we introduce \textbf{FineState-Bench}, a benchmark that evaluates whether an agent can correctly ground an instruction to the intended UI control and reach the exact target state. FineState-Bench comprises 2,209 instances across desktop, web, and mobile platforms, spanning four interaction families and 23 UI component types, with each instance explicitly specifying an exact target state for fine-grained state setting. We further propose \textit{FineState-Metrics}, a four-stage diagnostic pipeline with stage-wise success rates: Localization Success Rate (SR@Loc), Interaction Success Rate (SR@Int), Exact State Success Rate at Locate (ES-SR@Loc), and Exact State Success Rate at Interact (ES-SR@Int), and a plug-and-play \textit{Visual Diagnostic Assistant} (VDA) that generates a Description and a bounding-box Localization Hint to diagnose visual grounding reason via controlled w/ vs.\ w/o comparisons. On FineState-Bench, exact goal-state success remains low: ES-SR@Int peaks at 32.8% on Web and 22.8% on average across platforms. With VDA localization hints, Gemini-2.5-Flash gains +14.9 ES-SR@Int points, suggesting substantial headroom from improved visual grounding, yet overall accuracy is still insufficient for reliable fine-grained state-conditioned interaction \href{https://github.com/FengxianJi/FineState-Bench}{Github.} |
|
| The Effects of Visual Priming on Cooperative Behavior in Vision-Language Models | 2026-04-30 | ShowAs Vision-Language Models (VLMs) become increasingly integrated into decision-making systems, it is essential to understand how visual inputs influence their behavior. This paper investigates the effects of visual priming on VLMs' cooperative behavior using the Iterated Prisoner's Dilemma (IPD) as a test scenario. We examine whether exposure to images depicting behavioral concepts (kindness/helpfulness vs. aggressiveness/selfishness) and color-coded reward matrices alters VLM decision patterns. Experiments were conducted across multiple state-of-the-art VLMs. We further explore mitigation strategies including prompt modifications, Chain of Thought (CoT) reasoning, and visual token reduction. Results show that VLM behavior can be influenced by both image content and color cues, with varying susceptibility and mitigation effectiveness across models. These findings not only underscore the importance of robust evaluation frameworks for VLM deployment in visually rich and safety-critical environments, but also highlight how architectural and training differences among models may lead to distinct behavioral responses-an area worthy of further investigation. |
|
| Dynamic Cluster Data Sampling for Efficient and Long-Tail-Aware Vision-Language Pre-training | 2026-04-30 | ShowThe computational cost of training a vision-language model (VLM) can be reduced by sampling the training data. Previous work on efficient VLM pre-training has pointed to the importance of semantic data balance, adjusting the distribution of topics in the data to improve VLM accuracy. However, existing efficient pre-training approaches may disproportionately remove rare concepts from the training corpus. As a result, \emph{long-tail concepts} remain insufficiently represented in the training data and are not effectively captured during training. In this work, we introduce a \emph{dynamic cluster-based sampling approach (DynamiCS)} that downsamples large clusters of data and upsamples small ones. The approach is dynamic in that it applies sampling at each epoch. We first show the importance of dynamic sampling for VLM training. Then, we demonstrate the advantage of our cluster-scaling approach, which maintains the relative order of semantic clusters in the data and emphasizes the long-tail. This approach contrasts with current work, which focuses only on flattening the semantic distribution of the data. Our experiments show that DynamiCS reduces the computational cost of VLM training and provides a performance advantage for long-tail concepts. |
|
| OmniDrive-R1: Reinforcement-driven Interleaved Multi-modal Chain-of-Thought for Trustworthy Vision-Language Autonomous Driving | 2026-04-30 | ShowThe deployment of Vision-Language Models (VLMs) in safety-critical domains like autonomous driving (AD) is critically hindered by reliability failures, most notably object hallucination. This failure stems from their reliance on ungrounded, text-based Chain-of-Thought (CoT) reasoning. While existing multi-modal CoT approaches attempt mitigation, they suffer from two fundamental flaws: (1) decoupled perception and reasoning stages that prevent end-to-end joint optimization, and (2) reliance on expensive, dense localization labels. Thus we introduce OmniDrive-R1, an end-to-end VLM framework designed for autonomous driving, which unifies perception and reasoning through an interleaved Multi-modal Chain-of-Thought (iMCoT) mechanism. Our core innovation is an Reinforcement-driven visual grounding capability, enabling the model to autonomously direct its attention and "zoom in" on critical regions for fine-grained analysis. This capability is enabled by our pure two-stage reinforcement learning training pipeline and Clip-GRPO algorithm. Crucially, Clip-GRPO introduces an annotation-free, process-based grounding reward. This reward not only eliminates the need for dense labels but also circumvents the instability of external tool calls by enforcing real-time cross-modal consistency between the visual focus and the textual reasoning. Extensive experiments on DriveLMM-o1 demonstrate our model's significant improvements. Compared to the baseline Qwen2.5VL-7B, OmniDrive-R1 improves the overall reasoning score from 51.77% to 80.35%, and the final answer accuracy from 37.81% to 73.62%. |
|
| Iterative Multimodal Retrieval-Augmented Generation for Medical Question Answering | 2026-04-30 | ShowMedical retrieval-augmented generation (RAG) systems typically operate on text chunks extracted from biomedical literature, discarding the rich visual content (tables, figures, structured layouts) of original document pages. We propose MED-VRAG, an iterative multimodal RAG framework that retrieves and reasons over PMC document page images instead of OCR'd text. The system pairs ColQwen2.5 patch-level page embeddings with a sharded MapReduce LLM filter, scaling to ~350K pages while keeping Stage-1 retrieval under 30 ms via an offline coarse-to-fine index (C=8 centroids per page, ANN over centroids, exact two-way scoring on the top-R shortlist). A vision-language model (VLM) then iteratively refines its query and accumulates evidence in a memory bank across up to 3 reasoning rounds, with a single iteration costing ~15.9 s and the full three-round pipeline ~47.8 s on 4xA100. Across four medical QA benchmarks (MedQA, MedMCQA, PubMedQA, MMLU-Med), MEDVRAG reaches 78.6% average accuracy. Under controlled comparison with the same Qwen2.5-VL-32B backbone, retrieval contributes a +5.8 point gain over the no-retrieval baseline; we also note a +1.8 point edge over MedRAG + GPT-4 (76.8%), with the caveat that this is a cross-paper rather than head-to-head comparison. Ablations isolate +1.0 from page-image vs text-chunk retrieval, +1.5 from iteration, and +1.0 from the memory bank. |
|
| Auditing Frontier Vision-Language Models for Trustworthy Medical VQA: Grounding Failures, Format Collapse, and Domain Adaptation | 2026-04-30 | ShowDeploying vision-language models (VLMs) in clinical settings demands auditable behavior under realistic failure conditions, yet the failure landscape of frontier VLMs on specialized medical inputs is poorly characterized. We audit five recent frontier and grounding-aware VLMs (Gemini |
|
| Improving Calibration in Test-Time Prompt Tuning for Vision-Language Models via Data-Free Flatness-Aware Prompt Pretraining | 2026-04-30 | ShowTest-time prompt tuning (TPT) has emerged as a promising technique for enhancing the adaptability of vision-language models by optimizing textual prompts using unlabeled test data. However, prior studies have observed that TPT often produces poorly calibrated models, raising concerns about the reliability of their predictions. Recent works address this issue by incorporating additional regularization terms that constrain model outputs, which improve calibration but often degrade performance. In this work, we reveal that these regularization strategies implicitly encourage optimization toward flatter minima, and that the sharpness of the loss landscape around adapted prompts is a key factor governing calibration quality. Motivated by this observation, we introduce Flatness-aware Prompt Pretraining (FPP), a simple yet effective pretraining framework for TPT that initializes prompts within flatter regions of the loss landscape prior to adaptation. We show that simply replacing the initialization in existing TPT pipelines--without modifying any other components--is sufficient to improve both calibration and performance. Notably, FPP requires no labeled data and incurs no additional computational costs during test-time tuning, making it highly practical for real-world deployment. The code is available at: https://github.com/YonseiML/fpp. |
CVPR 2026 |
| ImagineNav++: Prompting Vision-Language Models as Embodied Navigator through Scene Imagination | 2026-04-30 | ShowVisual navigation is a fundamental capability for autonomous home-assistance robots, enabling long-horizon tasks such as object search. While recent methods have leveraged Large Language Models (LLMs) to incorporate commonsense reasoning and improve exploration efficiency, their planning remains constrained by textual representations, which cannot adequately capture spatial occupancy or scene geometry--critical factors for navigation decisions. We explore whether Vision-Language Models (VLMs) can achieve mapless visual navigation using only onboard RGB/RGB-D streams, unlocking their potential for spatial perception and planning. We achieve this through an imagination-powered navigation framework, ImagineNav++, which imagines future observation images from candidate robot views and translates navigation planning into a simple best-view image selection problem for VLMs. First, a future-view imagination module distills human navigation preferences to generate semantically meaningful viewpoints with high exploration potential. These imagined views then serve as visual prompts for the VLM to identify the most informative viewpoint. To maintain spatial consistency, we develop a selective foveation memory mechanism, which hierarchically integrates keyframe observations via a sparse-to-dense framework, constructing a compact yet comprehensive memory for long-term spatial reasoning. This approach transforms goal-oriented navigation into a series of tractable point-goal navigation tasks. Extensive experiments on open-vocabulary object and instance navigation benchmarks show that ImagineNav++ achieves SOTA performance in mapless settings, even surpassing most map-based methods, highlighting the importance of scene imagination and memory in VLM-based spatial reasoning. |
17 pa...17 pages, 10 figures. arXiv admin note: text overlap with arXiv:2410.09874 |
| WaferSAGE: Large Language Model-Powered Wafer Defect Analysis via Synthetic Data Generation and Rubric-Guided Reinforcement Learning | 2026-04-30 | ShowWe present WaferSAGE, a framework for wafer defect visual question answering using small vision-language models. To address data scarcity in semiconductor manufacturing, we propose a three-stage synthesis pipeline incorporating structured rubric generation for precise evaluation. Starting from limited labeled wafer maps, we employ clustering-based cleaning to filter label noise, then generate comprehensive defect descriptions using vision-language models, which are converted into structured evaluation rubrics criteria. These rubrics guide the synthesis of VQA pairs, ensuring coverage across defect type identification, spatial distribution, morphology, and root cause analysis. Our dual assessment framework aligns rule-based metrics with LLM-Judge scores via Bayesian optimization, enabling reliable automated evaluation. Through curriculum-based reinforcement learning with Group Sequence Policy Optimization (GSPO) and rubric-aligned rewards, our 4B-parameter Qwen3-VL model achieves a 6.493 LLM-Judge score, closely approaching Gemini-3-Flash (7.149) while enabling complete on-premise deployment. We demonstrate that small models with domain-specific training can surpass proprietary large models in specialized industrial visual understanding, offering a viable path for privacy-preserving, cost-effective deployment in semiconductor manufacturing. |
16 pa...16 pages, 3 figures, 8 tables |
| Test-Time Distillation for Continual Model Adaptation | 2026-04-30 | ShowDeep neural networks often suffer performance degradation upon deployment due to distribution shifts. Continual Test-Time Adaptation (CTTA) aims to address this issue in an unsupervised manner. However, existing methods that rely on self-supervision are prone to an inherent self-referential feedback loop that amplifies initial prediction errors, leading to model drift. We revisit this limitation and propose Test-Time Distillation (TTD), which reframes adaptation as a distillation process guided by a frozen Vision-Language Model (VLM) as an external signal. While promising, we find that direct distillation is fraught with two pitfalls: (1) the Generalist Trap, where the VLM's broad but non-specialized knowledge leads to suboptimal performance on specific tasks and shifts; and (2) the Entropy Bias, where naive model fusion techniques based on entropy fail due to the disparate calibration of heterogeneous models. These pitfalls highlight the need to build a robust supervisory signal and leverage it to guide the target model toward stable adaptation. Hence, we present CoDiRe, a Continual Distillation and Rectification framework for TTD. CoDiRe first constructs a robust blended teacher by dynamically fusing the predictions of the VLM and the target model. Critically, it circumvents the Entropy Bias by leveraging Maximum Softmax Probability (MSP) as a more reliable confidence metric for weighting each model's expertise. Then it applies an Optimal Transport-based rectification to further align predictions with the blended teacher, enabling continuous and stable adaptation. Extensive experiments show that CoDiRe outperforms state-of-the-art baselines, exceeding CoTTA by 10.55% with only 48% of its time cost on ImageNet-C. Project page is publicly available at https://github.com/walawalagoose/TTD. |
Accep...Accepted by CVPR 2026 Findings |
| CheXthought: A global multimodal dataset of clinical chain-of-thought reasoning and visual attention for chest X-ray interpretation | 2026-04-30 | ShowChest X-ray interpretation is one of the most frequently performed diagnostic tasks in medicine and a primary target for AI development, yet current vision-language models are primarily trained on datasets of paired images and reports, not the cognitive processes and visual attention that underlie clinical reasoning. Here, we present CheXthought, a global, multimodal resource containing 103,592 chain-of-thought reasoning traces and 6,609,082 synchronized visual attention annotations across 50,312 multi-read chest X-rays from 501 radiologists in 71 countries. Our analysis reveals clinical reasoning patterns in how experts deploy distinct visual search strategies, integrate clinical context, and communicate uncertainty. We demonstrate the clinical utility of CheXthought across four dimensions. First, CheXthought reasoning significantly outperforms state-of-the-art vision-language model chain-of-thought in factual accuracy and spatial grounding. Second, visual attention data used as an inference-time hint recovers missed findings and significantly reduces hallucinations. Third, vision-language models trained on CheXthought data achieve significantly stronger pathology classification, visual faithfulness, temporal reasoning and uncertainty communication. Fourth, leveraging CheXthought's multi-reader annotations, we predict both human-human and human-AI disagreement directly from an image, enabling transparent communication of case difficulty, uncertainty and model reliability. These findings establish CheXthought as a resource for advancing multimodal clinical reasoning and the development of more transparent, interpretable vision-language models. |
51 pa...51 pages, 7 figures, 10 tables |
| EdgeFM: Efficient Edge Inference for Vision-Language Models | 2026-04-30 | ShowVision-language models (VLMs) have demonstrated strong applicability in edge industrial applications, yet their deployment remains severely constrained by requirements for deterministic low latency and stable execution under resource limitations. Existing frameworks either rely on bloated general-purpose designs or force developers into opaque, hardware-specific closed-source ecosystems, leading to hardware lock-in limitation and poor cross-platform adaptability. Observing that modern AI agents can efficiently search and tune configurations to generate highly optimized low-level kernels for standard LLM operators, we propose EdgeFM, a lightweight, agent-driven VLM/LLM inference framework tailored for cross-platform industrial edge deployment. EdgeFM removes non-essential features to reduce single-request latency, and encapsulates agent-tuned kernel optimizations as a modular library of reusable skills. By allowing direct invocation of these skills rather than waiting for closed-source implementations, it effectively closes the performance gap long dominated by proprietary toolchains. The framework natively supports mainstream platforms including x86 and NVIDIA Orin SoCs, and represents the first end-to-end VLA deployment on the domestic Horizon Journey platform, enhancing cross-platform portability. In most cases, it yields clearly better inference performance than conventional vendor-specific toolchains, achieving up to 1.49 times speedup over TensorRT-Edge-LLM on the NVIDIA Orin platform. Experimental results show that EdgeFM delivers favorable end-to-end inference performance, providing an open-source, production-grade solution for diverse edge industrial scenarios. |
Techn...Technique Report version |
| RosettaSearch: Multi-Objective Inference-Time Search for Protein Sequence Design | 2026-04-30 | ShowWe introduce RosettaSearch, an inference-time multi-objective optimization approach for backbone conditioned protein sequence design. We use large language models (LLMs) as a generative optimizer within a search algorithm capable of controlled exploration and exploitation, using rewards computed from RosettaFold3, a structure prediction model, under a strict computational budget. In a large-scale evaluation, we apply RosettaSearch to 400 suboptimal sequences generated by LigandMPNN (a state-of-the-art model trained for protein sequence design), recovering high-fidelity designs that LigandMPNN's single-pass decoding fails to produce. RosettaSearch's designs show improvements in structural fidelity metrics ranging between 18% to 68%, translating to a 2.5x improvement in design success rate. We observe that these gains in success rate are robust when RosettaSearch-designed sequences are evaluated with an independent structure prediction oracle (Chai-1) and generalize across two distinct LLM families (o4-mini and Gemini-3), with performance scaling consistently with reasoning capability. We further demonstrate that RosettaSearch improves the sequence fidelity of ProteinMPNN designs for de novo backbones from the Dayhoff atlas, showing that the approach generalizes beyond native protein structures to computationally generated backbones. We also demonstrate a multi-modal extension of RosettaSearch with vision-language models, where images of predicted protein structures are used as feedback to incorporate structural context to guide protein sequence generation. To our knowledge, this is the first large-scale demonstration that LLMs can serve as effective generative optimizers for backbone-conditioned protein sequence design, yielding systematic gains without any model retraining. |
|
| Understanding Adversarial Transferability in Vision-Language Models for Autonomous Driving: A Cross-Architecture Analysis | 2026-04-30 | ShowVision-language models (VLMs) are increasingly used in autonomous driving because they combine visual perception with language-based reasoning, supporting more interpretable decision-making, yet their robustness to physical adversarial attacks, especially whether such attacks transfer across different VLM architectures, is not well understood and poses a practical risk when attackers do not know which model a vehicle uses. We address this gap with a systematic cross-architecture study of adversarial transferability in VLM-based driving, evaluating three representative architectures (Dolphins, OmniDrive, and LeapVAD) using physically realizable patches placed on roadside infrastructure in both crosswalk and highway scenarios. Our transfer-matrix evaluation shows high cross-architecture effectiveness, with transfer rates of 73-91% (mean TR = 0.815 for crosswalk and 0.833 for highway) and sustained frame-level manipulation over 64.7-79.4% of the critical decision window even when patches are not optimized for the target model. |
9 pag...9 pages, 2 figures. Accepted at SAE WCX 2026 |
| VeraRetouch: A Lightweight Fully Differentiable Framework for Multi-Task Reasoning Photo Retouching | 2026-04-30 | ShowReasoning photo retouching has gained significant traction, requiring models to analyze image defects, give reasoning processes, and execute precise retouching enhancements. However, existing approaches often rely on non-differentiable external software, creating optimization barriers and suffering from high parameter redundancy and limited generalization. To address these challenges, we propose VeraRetouch, a lightweight and fully differentiable framework for multi-task photo retouching. We employ a 0.5B Vision-Language Model (VLM) as the central intelligence to formulate retouching plans based on instructions and scene semantics. Furthermore, we develop a fully differentiable Retouch Renderer that replaces external tools, enabling direct end-to-end pixel-level training through decoupled control latents for lighting, global color, and specific color adjustments. To overcome data scarcity, we introduce AetherRetouch-1M+, the first million-scale dataset for professional retouching, constructed via a new inverse degradation workflow. Furthermore, we propose DAPO-AE, a reinforcement learning post-training strategy that enhances autonomous aesthetic cognition. Extensive experiments demonstrate that VeraRetouch achieves state-of-the-art performance across multiple benchmarks while maintaining a significantly smaller footprint, enabling mobile deployment. Our code and models are publicly available at https://github.com/OpenVeraTeam/VeraRetouch. |
|
| CasLayout: Cascaded 3D Layout Diffusion for Indoor Scene Synthesis with Implicit Relation Modeling | 2026-04-30 | ShowSynthesizing realistic 3D indoor scenes remains challenging due to data scarcity and the difficulty of simultaneously enforcing global architectural constraints and local semantic consistency. Existing approaches often overlook structural boundaries or rely on fully connected relation graphs that introduce redundant generation errors. Inspired by human design cognition, we present CasLayout, a cascaded diffusion framework that decomposes the joint scene generation task into four conditional sub-stages with explicit physical and semantic roles: (1) predicting furniture quantity and categories, (2) refining object sizes and feature embeddings, (3) modeling spatial relationships in a latent space, and (4) generating Oriented Bounding Boxes (OBBs). This decoupled architecture reduces data requirements and enables flexible integration of Large Language Models (LLMs) and Vision Language Models (VLMs) for zero-shot tasks such as image-to-scene generation. To maintain physical validity within complex floor plans, we explicitly model building elements (e.g., walls, doors, and windows) as conditional constraints. Furthermore, to address the high entropy of dense relation graphs, we introduce a sparse relation graph formulation aligned with human spatial descriptions. By encoding these sparse graphs into a compact latent space using a bidirectional Variational Autoencoder (VAE), the proposed framework provides enhanced relational controllability, allowing generated layouts to better respect functional organization. Experiments demonstrate that CasLayout achieves state-of-the-art performance in fidelity and diversity while enabling improved controllability in practical applications. |
SIGGA...SIGGARPH 2026 (Journal Track), Code: https://github.com/YingruiWoo/CasLayout |
| Do World Action Models Generalize Better than VLAs? A Robustness Study | 2026-04-30 | ShowRobot action planning in the real world is challenging as it requires not only understanding the current state of the environment but also predicting how it will evolve in response to actions. Vision-language-action (VLA), which repurpose large-scale vision-language models for robot action generation using action experts, have achieved notable success across a variety of robotic tasks. Nevertheless, their performance remains constrained by the scope of their training data, exhibiting limited generalization to unseen scenarios and vulnerability to diverse contextual perturbations. More recently, world models have been revisited as an alternative to VLAs. These models, referred to as world action models (WAMs), are built upon world models that are trained on large corpora of video data to predict future states. With minor adaptations, their latent representation can be decoded into robot actions. It has been suggested that their explicit dynamic prediction capacity, combined with spatiotemporal priors acquired from web-scale video pretraining, enables WAMs to generalize more effectively than VLAs. In this paper, we conduct a comparative study of prominent state-of-the-art VLA policies and recently released WAMs. We evaluate their performance on the LIBERO-Plus and RoboTwin 2.0-Plus benchmarks under various visual and language perturbations. Our results show that WAMs achieve strong robustness, with LingBot-VA reaching 74.2% success rate on RoboTwin 2.0-Plus and Cosmos-Policy achieving 82.2% on LIBERO-Plus. While VLAs such as |
|
| OpAgent: Operator Agent for Web Navigation | 2026-04-30 | ShowTo fulfill user instructions, autonomous web agents must contend with the inherent complexity and volatile nature of real-world websites. Conventional paradigms predominantly rely on Supervised Fine-Tuning (SFT) or Offline Reinforcement Learning (RL) using static datasets. However, these methods suffer from severe distributional shifts, as offline trajectories fail to capture the stochastic state transitions and real-time feedback of unconstrained wide web environments. In this paper, we propose a robust Online Reinforcement Learning WebAgent, designed to optimize its policy through direct, iterative interactions with unconstrained wide websites. Our approach comprises three core innovations: 1) Hierarchical Multi-Task Fine-tuning: We curate a comprehensive mixture of datasets categorized by functional primitives -- Planning, Acting, and Grounding -- establishing a Vision-Language Model (VLM) with strong instruction-following capabilities for Web GUI tasks. 2) Online Agentic RL in the Wild: We develop an online interaction environment and fine-tune the VLM using a specialized RL pipeline. We introduce a Hybrid Reward Mechanism that combines a ground-truth-agnostic WebJudge for holistic outcome assessment with a Rule-based Decision Tree (RDT) for progress reward. This system effectively mitigates the credit assignment challenge in long-horizon navigation. Notably, our RL-enhanced model achieves a 38.1% success rate (pass@5) on WebArena, outperforming all existing monolithic baselines. 3) Operator Agent: We introduce a modular agentic framework, namely \textbf{OpAgent}, orchestrating a Planner, Grounder, Reflector, and Summarizer. This synergy enables robust error recovery and self-correction, elevating the agent's performance to a new State-of-the-Art (SOTA) success rate of \textbf{71.6%}. |
|
| AutoVDC: Automated Vision Data Cleaning Using Vision-Language Models | 2026-04-30 | ShowTraining of autonomous driving systems requires extensive datasets with precise annotations to attain robust performance. Human annotations suffer from imperfections, and multiple iterations are often needed to produce high-quality datasets. However, manually reviewing large datasets is laborious and expensive. In this paper, we introduce AutoVDC (Automated Vision Data Cleaning) framework and investigate the utilization of Vision-Language Models (VLMs) to automatically identify erroneous annotations in vision datasets, thereby enabling users to eliminate these errors and enhance data quality. We validate our approach using the KITTI and nuImages datasets, which contain object detection benchmarks for autonomous driving. To test the effectiveness of AutoVDC, we create dataset variants with intentionally injected erroneous annotations and observe the error detection rate of our approach. Additionally, we compare the detection rates using different VLMs and explore the impact of VLM fine-tuning on our pipeline. The results demonstrate our method's high performance in error detection and data cleaning experiments, indicating its potential to significantly improve the reliability and accuracy of large-scale production datasets in autonomous driving. |
Accep...Accepted to IV 2026 Drive-X Foundation Models for Autonomous Driving (Oral presentation) |
| Vision-Language Models Mistake Head Orientation for Gaze Direction: Nonverbal Conversation Cues | 2026-04-29 | ShowWhere someone looks is a nonverbal communication cue that children and adults readily use. How well can Vision-Language Models (VLMs) infer gaze targets? To construct evaluation stimuli, we captured 1,360 real-world photos of scenes in which a person gazes at one of several objects on a table. Importantly, we also controlled the gazer's head orientation: sometimes it was directed toward the gaze target, sometimes toward a distractor object, and sometimes left unconstrained. We found a substantial performance gap between VLMs and humans, ruled out alternative explanations such as resolution and object-naming skills, and identified the main reason for the gap as VLMs inferring gaze direction using head orientation rather than eye appearance. Such a bias is likely due to data rather than architecture, as suggested by a proof-of-concept experiment finetuning a transformer-based vision model. Future work should investigate whether these findings hold broadly across various deep learning methods trained on existing data, and whether better data mitigates this problem for all architectures. Pinpointing the reason sets the stage for technologies that can interpret gaze targets to have more efficient interactions with humans. |
Accep...Accepted by ACL 2026. Project page at https://zoryzhang.github.io/gaze/ |
| TeD-Loc: Text Distillation for Weakly Supervised Object Localization | 2026-04-29 | ShowWeakly supervised object localization (WSOL) models are trained using only image-level class labels. They can predict both the object class and spatial regions corresponding to the object, without requiring explicit bounding box annotations. Given their reliance on classification objectives, traditional WSOL methods, like class activation mapping, tend to focus on the most discriminative object regions, often missing the full spatial extent. Although vision-language models such as CLIP encode rich semantic priors, they are not directly suited for WSOL because global text and class-token embeddings are not explicitly aligned with local patch embeddings, making patch-level localization difficult without additional mechanisms. Recent methods such as GenPrompt address this limitation, but at the cost of increased complexity, as they rely on conditional denoising and elaborate prompt-learning strategies. We propose Text Distillation for Localization (TeD-Loc), which transfers knowledge from CLIP text embeddings to patch embeddings through contrastive alignment, thereby enabling patch-level foreground/background localization. A localization-guided classification module is also introduced that uses localization scores to aggregate foreground patch embeddings for joint classification and localization in a single model. In addition, a QR-based orthogonalization of class text embeddings is applied before distillation to improve discrimination for semantically similar classes. Extensive experiments show that TeD-Loc improves Top-1 Loc by ~5% on CUB and ILSVRC, and PxAP by ~31% on histopathology benchmarks, while achieving more efficient inference than GenPrompt. |
|
| InterPartAbility: Text-Guided Part Matching for Interpretable Person Re-Identification | 2026-04-29 | ShowText-to-image person re-identification (TI-ReID) relies on natural-language text description to retrieve top matching individuals from a large gallery of images. While recent large vision-language models (VLMs) achieve strong retrieval performance, their decisions remain largely uninterpretable. Existing interpretability approaches in TI-ReID rely solely on slot-attention to highlight attended regions, but fail to reliably bind visual regions to semantically meaningful concepts, limiting explanations to qualitative visualizations over a restricted vocabulary. This paper introduces InterPartAbility, an interpretable TI-ReID method that performs explicit part-wise matching and enables phrase-region grounding. A new open-vocabulary, lightweight supervision, patch-phrase interaction module (PPIM) is proposed to train a standard TI-ReID model with concept-level guidance. Concept-based part phrases provide evidence that encourages the model to attend to corresponding image regions. InterPartAbility further constrains CLIP ViT self-attention to produce spatially concentrated patch activations aligned with each part-level phrase, yielding grounded explanation maps. A quantitative interpretability protocol for TI-ReID is introduced by adapting perturbation-based evaluation metrics, including counterfactual region masking that measures retrieval degradation when top-ranked explanatory regions are removed. Empirical results\footnote{Our code is included in the supplementary materials and will be made public.} on challenging benchmarks like CUHK-PEDES and ICFG-PEDES show that InterPartAbility achieves state-of-the-art (SOTA) interpretability performance under these metrics, while sustaining competitive retrieval accuracy. |
|
| Context Matters: Peer-Aware Student Behavioral Engagement Measurement via VLM Action Parsing and LLM Sequence Classification | 2026-04-29 | ShowUnderstanding student behavior in the classroom is essential to improve both pedagogical quality and student engagement. Existing methods for predicting student engagement typically require substantial annotated data to model the diversity of student behaviors, yet privacy concerns often restrict researchers to their own proprietary datasets. Moreover, the classroom context, represented in peers' actions, is ignored. To address the aforementioned limitation, we propose a novel three-stage framework for video-based student engagement measurement. First, we explore the few-shot adaptation of the vision-language model for student action recognition, which is fine-tuned to distinguish among action categories with a few training samples. Second, to handle continuous and unpredictable student actions, we utilize the sliding temporal window technique to divide each student's 2-minute-long video into non-overlapping segments. Each segment is assigned an action category via the fine-tuned VLM model, generating a sequence of action predictions. Finally, we leverage the large language model to classify this entire sequence of actions, together with the classroom context, as belonging to an engaged or disengaged student. The experimental results demonstrate the effectiveness of the proposed approach in identifying student engagement. The source code will be available at https://github.com/ahmed-nady/context_aware_student_engagement. |
accep...accepted to the Computer Vision for Education (CV4Edu) workshop, CVPR 2026 |
| Lost in Space? Vision-Language Models Struggle with Relative Camera Pose Estimation | 2026-04-29 | ShowWe study whether vision-language models (VLMs) can solve relative camera pose estimation (RCPE) from image pairs, a direct test of multi-view spatial reasoning. We cast RCPE as a discrete verbal classification task and introduce \texttt{VRRPI-Bench}, built from real RGB-D frames with object-centric camera motion, and \texttt{VRRPI-Diag}, which isolates individual motion degrees of freedom. Humans (0.91) and specialized geometric pipelines such as LoFTR (0.99) solve the task reliably, yet the best VLM reaches only 0.66 and most others remain near random. Our analyses show that this gap is not basic spatial competence: strong VLMs are near ceiling on single-image benchmarks, but most remain near random once reasoning must span views. They are unstable under source-target reversal (best 59.7% consistency) and remain weak even in simplified single-DoF settings, especially on optical-axis motions such as roll and depth translation (GPT-5: 0.46 on roll). These failures are useful: they localize concrete missing capabilities, namely cross-view correspondence, view-consistent reasoning, and projective camera-motion understanding, making RCPE a targeted diagnostic for improving multi-view spatial reasoning in VLMs. |
|
| World2VLM: Distilling World Model Imagination into VLMs for Dynamic Spatial Reasoning | 2026-04-29 | ShowVision-language models (VLMs) have shown strong performance on static visual understanding, yet they still struggle with dynamic spatial reasoning that requires imagining how scenes evolve under egocentric motion. Recent efforts address this limitation either by scaling spatial supervision with synthetic data or by coupling VLMs with world models at inference time. However, the former often lacks explicit modeling of motion-conditioned state transitions, while the latter incurs substantial computational overhead. In this work, we propose World2VLM, a training framework that distills spatial imagination from a generative world model into a vision-language model. Given an initial observation and a parameterized camera trajectory, we use a view-consistent world model to synthesize geometrically aligned future views and derive structured supervision for both forward (action-to-outcome) and inverse (outcome-to-action) spatial reasoning. We post-train the VLM with a two-stage recipe on a compact dataset generated by this pipeline and evaluate it on multiple spatial reasoning benchmarks. World2VLM delivers consistent improvements over the base model across diverse benchmarks, including SAT-Real, SAT-Synthesized, VSI-Bench, and MindCube. It also outperforms the test-time world-model-coupled methods while eliminating the need for expensive inference-time generation. Our results suggest that world models can serve not only as inference-time tools, but also as effective training-time teachers, enabling VLMs to internalize spatial imagination in a scalable and efficient manner. |
The c...The code is available at https://github.com/WanyueZhang-ai/World2VLM. The dataset is available at https://huggingface.co/datasets/WanyueZhang/World2VLM |
| Value-Guided Iterative Refinement and the DIQ-H Benchmark for Evaluating VLM Robustness | 2026-04-29 | ShowVision-Language Models (VLMs) are essential for embodied AI and safety-critical applications, such as robotics and autonomous systems. However, existing benchmarks primarily focus on static or curated visual inputs, neglecting the challenges posed by adversarial conditions, value misalignment, and error propagation in continuous deployment. Current benchmarks either overlook the impact of real-world perturbations, or fail to account for the cumulative effect of inconsistent reasoning over time. To address these gaps, we introduce the Degraded Image Quality Leading to Hallucinations (DIQ-H) benchmark, the first to evaluate VLMs under adversarial visual conditions in continuous sequences. DIQ-H simulates real-world stressors including motion blur, sensor noise, and compression artifacts, and measures how these corruptions lead to persistent errors and misaligned outputs across time. The benchmark explicitly models error propagation and its long-term value consistency. To enhance scalability and reduce costs for safety-critical evaluation, we propose the Value-Guided Iterative Refinement (VIR) framework, which automates the generation of high-quality, ethically aligned ground truth annotations. VGIR leverages lightweight VLMs to detect and refine value misalignment, improving accuracy from 72.2% to 83.3%, representing a 15.3% relative improvement. The DIQ-H benchmark and VGIR framework provide a robust platform for embodied AI safety assessment, revealing vulnerabilities in error recovery, ethical consistency, and temporal value alignment. |
|
| Walk With Me: Long-Horizon Social Navigation for Human-Centric Outdoor Assistance | 2026-04-29 | ShowAssisting humans in open-world outdoor environments requires robots to translate high-level natural-language intentions into safe, long-horizon, and socially compliant navigation behavior. Existing map-based methods rely on costly pre-built HD maps, while learning-based policies are mostly limited to indoor and short-horizon settings. To bridge this gap, we propose Walk with Me, a map-free framework for long-horizon social navigation from high-level human instructions. Walk with Me leverages GPS context and lightweight candidate points-of-interest from a public map API for semantic destination grounding and waypoint proposal. A High-Level Vision-Language Model grounds abstract instructions into concrete destinations and plans coarse waypoint sequences. During execution, an observation-aware routing mechanism determines whether the Low-Level Vision-Language-Action policy can handle the current situation or whether explicit safety reasoning from the High-Level VLM is needed. Routine segments are executed by the Low-Level VLA, while complex situations such as crowded crossings trigger high-level reasoning and stop-and-wait behavior when unsafe. By combining semantic intent grounding, map-free long-horizon planning, safety-aware reasoning, and low-level action generation, Walk with Me enables practical outdoor social navigation for human-centric assistance. |
|
| Contrastive Semantic Projection: Faithful Neuron Labeling with Contrastive Examples | 2026-04-29 | ShowNeuron labeling assigns textual descriptions to internal units of deep networks. Existing approaches typically rely on highly activating examples, often yielding broad or misleading labels by focusing on dominant but incidental visual factors. Prior work such as FALCON introduced contrastive examples -- inputs that are semantically similar to activating examples but elicit low activations -- to sharpen explanations, but it primarily addresses subspace-level interpretability rather than scalable neuron-level labeling. We revisit contrastive explanations for neuron-level labeling in two stages: (1) candidate label generation with vision language models (VLMs) and (2) label assignment with CLIP-like encoders. First, we show that providing contrastive image sets to VLMs yields candidate labels that are more specific and more faithful. Second, we introduce Contrastive Semantic Projection (CSP), an extension of SemanticLens that incorporates contrastive examples directly into its CLIP-based scoring and selection pipeline. Across extensive experiments and a case study on melanoma detection, contrastive labeling improves both faithfulness and semantic granularity over state-of-the-art baselines. Our results demonstrate that contrastive examples are a simple yet powerful and currently underutilized component of neuron labeling and analysis pipelines. |
|
| SynSur: An end-to-end generative pipeline for synthetic industrial surface defect generation and detection | 2026-04-29 | ShowThe bottleneck in learning-based industrial defect detection is often limited not by model capacity, but by the scarcity of labeled defect data: defects are rare, annotations are expensive, and collecting balanced training sets is slow. We present an end-to-end pipeline for synthetic defect generation and annotation, combining Vision-Language-Model-based prompts, LoRA-adapted diffusion, mask-guided inpainting, and sample filtering with automatic label derivation, and demonstrates the potential of real data with realistic synthetic samples to overcome data scarcity. The evaluation is conducted on, a challenging dataset of pitting defects on ball screw drives, and then on a subset of the Mobile phone screen surface defect segmentation dataset (MSD) dataset to test cross-domain transfer. Beyond downstream detector performance, we analyze key stages of the pipeline, including prompt construction, LoRA selection, and sample filtering with DreamSim and CLIPScore, to understand which synthetic samples are both realistic and useful. Experiments with YOLOv26, YOLOX, and LW-DETR show that synthetic-only training does not replace real data. When combined with real data, synthetic defects can preserve performance and yield modest gains in selected BSData training regimes. The MSD transfer study shows that the overall pipeline structure carries over to a second industrial inspection domain, while also highlighting the importance of domain-specific adaptation and annotation-quality control. Overall, the paper provides an end-to-end assessment of diffusion-based industrial defect synthesis and shows that its strongest value lies in strengthening scarce real datasets rather than substituting for them. |
|
| Prox-E: Fine-Grained 3D Shape Editing via Primitive-Based Abstractions | 2026-04-29 | ShowText-based 2D image editing models have recently reached an impressive level of maturity, motivating a growing body of work that heavily depends on these models to drive 3D edits. While effective for appearance-based modifications, such 2D-centric 3D editing pipelines often struggle with fine-grained 3D editing, where localized structural changes must be applied while strictly preserving an object's overall identity. To address this limitation, we propose Prox-E, a training-free framework that enables fine-grained 3D control through an explicit, primitive-based geometric abstraction. Our framework first abstracts an input 3D shape into a compact set of geometric primitives. A pretrained vision-language model (VLM) then edits this abstraction to specify primitive-level changes. These structural edits are subsequently used to guide a 3D generative model, enabling fine-grained, localized modifications while preserving unchanged regions of the original shape. Through extensive experiments, we demonstrate that our method consistently balances identity preservation, shape quality, and instruction fidelity more effectively than various existing approaches, including 2D-based 3D editors and training-based methods. |
Accep...Accepted to SIGGRAPH 2026. Project page: https://etaisella.github.io/Prox-E/ |
| ViTaPEs: Visuotactile Position Encodings for Cross-Modal Alignment in Multimodal Transformers | 2026-04-29 | ShowTactile sensing provides local essential information that is complementary to visual perception, such as texture, compliance, and force. Despite recent advances in visuotactile representation learning, challenges remain in fusing these modalities and generalizing across tasks and environments without heavy reliance on pre-trained vision-language models. Moreover, existing methods do not study positional encodings, thereby overlooking the multi-stage spatial reasoning needed to capture fine-grained visuotactile correlations. We introduce ViTaPEs, a transformer-based architecture for learning task-agnostic visuotactile representations from paired vision and tactile inputs. Our key idea is a two-stage positional injection: local (modality-specific) positional encodings are added within each stream, and a global positional encoding is added on the joint token sequence immediately before attention, providing a shared positional vocabulary at the stage where cross-modal interaction occurs. We make the positional injection points explicit and conduct controlled ablations that isolate their effect before a token-wise nonlinearity versus immediately before self-attention. Experiments on multiple large-scale real-world datasets show that ViTaPEs not only surpasses state-of-the-art baselines across various recognition tasks but also demonstrates zero-shot generalization to unseen, out-of-domain scenarios. We further demonstrate the transfer-learning strength of \emph{ViTaPEs} in a robotic grasping task, where it outperforms state-of-the-art baselines in predicting grasp success. Project page: https://sites.google.com/view/vitapes |
|
| Progressive Semantic Communication for Efficient Edge-Cloud Vision-Language Models | 2026-04-29 | ShowDeploying Vision-Language Models (VLMs) on edge devices remains challenging due to their substantial computational and memory demands, which exceed the capabilities of resource-constrained embedded platforms. Conversely, fully offloading inference to the cloud is often impractical in bandwidth-limited environments, where transmitting raw visual data introduces substantial latency overhead. While recent edge-cloud collaborative architectures attempt to partition VLM workloads across devices, they typically rely on transmitting fixed-size representations, lacking adaptability to dynamic network conditions and failing to fully exploit semantic redundancy. In this paper, we propose a progressive semantic communication framework for edge-cloud VLM inference, using a Meta AutoEncoder that compresses visual tokens into adaptive, progressively refinable representations, enabling plug-and-play deployment with off-the-shelf VLMs without additional fine-tuning. This design allows flexible transmission at different information levels, providing a controllable trade-off between communication cost and semantic fidelity. We implement a full end-to-end edge-cloud system comprising an embedded NXP i.MX95 platform and a GPU server, communicating over bandwidth-constrained networks. Experimental results show that, at 1 Mbps uplink, the proposed progressive scheme significantly reduces network latency compared to full-edge and full-cloud solutions, while maintaining high semantic consistency even under high compression. The implementation code will be released upon publication at https://github.com/open-ep/ProSemComVLM. |
Under...Under review. Extended version with additional figures and appendices |
| A Multistage Extraction Pipeline for Long Scanned Financial Documents: An Empirical Study in Industrial KYC Workflows | 2026-04-29 | ShowStructured information extraction from long, multilingual scanned financial documents is a core requirement in industrial KYC and compliance workflows. These documents are typically non machine readable, noisy, and visually heterogeneous. They usually span dozens of pages while containing only sparse task relevant information. Although recent vision-language models achieve strong benchmark performance, directly applying them end to end to full financial reports often leads to unreliable extraction under real world conditions. We present a multistage extraction framework that integrates image preprocessing, multilingual OCR, hybrid page-level retrieval, and compact VLM-based structured extraction. The design separates page localization from multimodal reasoning, enabling more accurate extraction from complex multipage documents. We evaluated the framework on 120 production KYC documents comprising about 3000 multilingual scanned pages. Across multiple OCR-VLM combinations, the proposed pipeline consistently outperforms direct PDF-to-VLM baselines, improving field-level accuracy by up to 31.9 percentage points. The best configuration, PaddleOCR with MiniCPM2.6, achieves 87.27 percent accuracy. Ablation studies show that page-level retrieval is the dominant factor in performance improvements, particularly for complex financial statements and non-English documents. |
|
| Delineating Knowledge Boundaries for Honest Large Vision-Language Models | 2026-04-29 | ShowLarge Vision-Language Models (VLMs) have achieved remarkable multimodal performance yet remain prone to factual hallucinations, particularly in long-tail or specialized domains. Moreover, current models exhibit a weak capacity to refuse queries that exceed their parametric knowledge. In this paper, we propose a systematic framework to enhance the refusal capability of VLMs when facing such unknown questions. We first curate a model-specific "Visual-Idk" (Visual-I don't know) dataset, leveraging multi-sample consistency probing to distinguish between known and unknown facts. We then align the model using supervised fine-tuning followed by preference-aware optimization (e.g., DPO, ORPO) to effectively delineate its knowledge boundaries. Results on the Visual-Idk dataset show our method improves the Truthful Rate from 57.9% to 67.3%. Additionally, internal probing also demonstrates that the model genuinely recognizes its boundaries instead of just memorizing refusal patterns. Our framework further generalizes to out-of-distribution medical and perceptual domains, providing a robust path toward more trustworthy and prudent visual assistants. |
|
| Topology-Aware Representation Alignment for Semi-Supervised Vision-Language Learning | 2026-04-29 | ShowVision-language models have shown strong performance, but they often generalize poorly to specialized domains. While semi-supervised vision-language learning mitigates this limitation by leveraging a small set of labeled image-text pairs together with abundant unlabeled images, existing methods remain fundamentally pairwise and fail to model the global structure of multimodal representation manifolds. Existing topology-based alignment methods rely on persistence diagram matching, which neither guarantees geometric alignment nor utilizes the image-text pairing information central to vision-language learning. We propose Topology-Aware Multimodal Representation Alignment (ToMA), a framework that uses persistent homology to identify topologically salient edges and aligns them across modalities through available cross-modal correspondences. ToMA leverages both H_0-death edges and lightweight H_1-birth edges, allowing it to capture both connectivity and cycle structure without constructing 2-simplices. Experiments show that ToMA yields stable gains, with clear improvements on remote sensing and modest but consistent benefits on fashion retrieval. Additional analysis shows that ToMA is more stable than alternative topology-based objectives and that lightweight H_1-birth edges provide useful higher-order structural signals. |
30 pa...30 pages, 10 figures, 24 tables |
| VLM Judges Can Rank but Cannot Score: Task-Dependent Uncertainty in Multimodal Evaluation | 2026-04-29 | ShowVision-language models (VLMs) are increasingly used as automated judges for multimodal systems, yet their scores provide no indication of reliability. We study this problem through conformal prediction, a distribution-free framework that converts a judge's point score into a calibrated prediction interval using only score-token log-probabilities, with no retraining. We present the first systematic analysis of conformal prediction for VLM-as-a-Judge across 3 judges and 14 visual task categories. Our results show that evaluation uncertainty is strongly task-dependent: intervals cover ~40% of the score range for aesthetics and natural images but expand to ~70% for chart and mathematical reasoning, yielding a quantitative reliability map for multimodal evaluation. We further identify a failure mode not captured by standard evaluation metrics, ranking-scoring decoupling, where judges achieve high ranking correlation while producing wide, uninformative intervals, correctly ordering responses but failing to assign reliable absolute scores. Finally, we show that interval width is driven primarily by task difficulty and annotation quality, i.e., the same judge and method yield 4.5x narrower intervals on a clean, multi-annotator captioning benchmark. Code: https://github.com/divake/VLM-Judge-Uncertainty |
|
| UIGaze: How Closely Can VLMs Approximate Human Visual Attention on User Interfaces? | 2026-04-29 | ShowVision Language Models (VLMs) have demonstrated strong capabilities in understanding visual content, yet their ability to predict where humans look on user interfaces remains unexplored. We present UIGaze, a study investigating how closely VLMs can approximate human visual attention on user interfaces using real eye-tracking data. Using the UEyes dataset - comprising 1,980 UI screenshots across four categories (webpage, desktop, mobile, poster) with eye-tracking data from 62 participants - we evaluate nine state-of-the-art VLMs through a zero-shot coordinate prediction pipeline. Each model generates gaze point coordinates that are converted into saliency maps via Gaussian blurring and compared against ground truth using CC, SIM, and KL divergence. Our experiments (1,980 images x 9 models x 3 runs x 3 durations) reveal that VLMs achieve moderate alignment with human gaze patterns, with the degree of alignment varying significantly across UI types and improving with longer viewing durations - suggesting VLMs capture exploratory gaze patterns rather than initial fixations. All code, predictions, and evaluation results are publicly available. |
6 pag...6 pages, 4 tables, 1 figure |
| Efficient, VRAM-Constrained xLM Inference on Clients | 2026-04-29 | ShowTo usher in the next round of client AI innovation, there is an urgent need to enable efficient, lossless inference of high-accuracy large language models (LLMs) and vision language models (VLMs), jointly referred to as xLMs, on client systems. To address this, we present pipelined sharding, a novel, benchmark-profile-guided CPU-GPU hybrid scheduling technique to achieve efficient, VRAM-constrained inference for both dense and mixture-of-experts (MoE) LLMs. Using a combination of model sharding at the sub-layer level, CPU offloading, pipelined copy-compute, and prioritized tensor placement in VRAM, it optimizes both time-to-first-token (TTFT) and tokens per second (TPS) metrics, while flexibly adapting to system and inference conditions. For efficient, high-accuracy VLM inference, we combine pipelined sharding with a llama$.$cpp implementation of three well-understood prior ideas (jointly called VLMOpt), namely, vision tensor CPU offloading, flash attention, and vision and language model VRAM overlap avoidance. These enhancements are targeted at improving client xLM inference in future releases of two important NVIDIA products - the In-Game Inferencing software development kit (IGI SDK) and the Cosmos-Reason1 (CR1) physical AI reasoning VLM. Highlights from our rigorous evaluation spanning multiple models and client systems include: for interactive use, TTFT improves by up to 6.7x and TPS by up to 30x for LLMs, and CR1 inference's VRAM demand is down by 10x, while in batched mode, throughput improves by up to 8.2x, all compared to their respective aggressive baselines. This paper is accepted at the 9th MLSys Conference (Industry Track), 2026. Code and artifact available at: https://github.com/deepshnv/pipeshard-mlsys26-ae |
Accep...Accepted at MLSys 2026 (Industry Track). 17 pages, 7 figures, 9 tables. Code and artifacts available at: https://github.com/deepshnv/pipeshard-mlsys26-ae |
| FLARE: Fully Integration of Vision-Language Representations for Deep Cross-Modal Understanding | 2026-04-29 | ShowWe introduce FLARE, a family of vision language models (VLMs) with a fully vision-language alignment and integration paradigm. Unlike existing approaches that rely on single MLP projectors for modality alignment and defer cross-modal interaction to LLM decoding, FLARE achieves deep, dynamic integration throughout the pipeline. Our key contributions include: (1) Text-Guided Vision Encoding that incorporates textual information during vision encoding to achieve pixel-level alignment; (2) Context-Aware Alignment Decoding that aggregates visual features conditioned on textual context during decoding for query-level integration; (3) Dual-Semantic Mapping Loss to supervise feature mapping from both modalities and enable modality-level bridging; and (4) Text-Driven VQA Synthesis that leverages high-quality text to generate VQA pairs and synthesize corresponding images, enabling data-level optimization. We train FLARE at 3B and 8B scales under both fixed and dynamic resolution settings, demonstrating that our full-modality alignment significantly outperforms existing methods while maintaining strong generalizability. FLARE 3B surpasses Cambrian-1 8B and Florence-VL 8B using only 630 vision tokens. Ablation studies reveal that FLARE achieves superior performance over existing methods with minimal computational cost. Even without dynamic resolution, FLARE outperforms LLaVA-NeXT, validating the effectiveness of our approach. We release our code, model weights, and dataset in https://github.com/starriver030515/FLARE. |
|
| ChartVerse: Scaling Chart Reasoning via Reliable Programmatic Synthesis from Scratch | 2026-04-29 | ShowChart reasoning is a critical capability for Vision Language Models (VLMs). However, the development of open-source models is severely hindered by the lack of high-quality training data. Existing datasets suffer from a dual challenge: synthetic charts are often simplistic and repetitive, while the associated QA pairs are prone to hallucinations and lack the reasoning depth required for complex tasks. To bridge this gap, we propose ChartVerse, a scalable framework designed to synthesize complex charts and reliable reasoning data from scratch. (1) To address the bottleneck of simple patterns, we first introduce Rollout Posterior Entropy (RPE), a novel metric that quantifies chart complexity. Guided by RPE, we develop complexity-aware chart coder to autonomously synthesize diverse, high-complexity charts via executable programs. (2) To guarantee reasoning rigor, we develop truth-anchored inverse QA synthesis. Diverging from standard generation, we adopt an answer-first paradigm: we extract deterministic answers directly from the source code, generate questions conditional on these anchors, and enforce strict consistency verification. To further elevate difficulty and reasoning depth, we filter samples based on model fail-rate and distill high-quality Chain-of-Thought (CoT) reasoning. We curate ChartVerse-SFT-600K and ChartVerse-RL-40K using Qwen3-VL-30B-A3B-Thinking as the teacher. Experimental results demonstrate that ChartVerse-8B achieves state-of-the-art performance, notably surpassing its teacher and rivaling the stronger Qwen3-VL-32B-Thinking. We release our code, model weights, and datasets in https://chartverse.github.io. |
29 pages |
| VLN-Cache: Enabling Token Caching for VLN Models with Visual/Semantic Dynamics Awareness | 2026-04-29 | ShowVision-and-Language Navigation (VLN) increasingly relies on large vision-language models, but their inference cost conflicts with real-time deployment. Token caching is a promising training-free strategy that avoids redundant computation by reusing stable visual tokens across frames. However, existing methods assume a static camera and fixed semantic focus, assumptions that VLN fundamentally violates. We identify two failure modes: (1) visual dynamics, where viewpoint shift displaces token positions across frames, causing position-wise matching to pair misaligned content; (2) semantic dynamics, where token relevance shifts across task stages as navigation progresses, making cached states stale. We propose VLN-Cache, a visual-dynamic-aware and semantic-dynamic-aware caching framework that introduces view-aligned remapping to recover geometric correspondences and a task-relevance saliency filter to veto reuse at semantic transitions. A layer-adaptive entropy policy further balances the per-layer reuse budget. Experiments on the R2R-CE simulation benchmark show up to 1.52x speedup while maintaining competitive navigation success rates. |
|
| CoFL: Continuous Flow Fields for Language-Conditioned Navigation | 2026-04-29 | ShowExisting language-conditioned navigation systems typically rely on modular pipelines or trajectory generators, but the latter use each scene--instruction annotation mainly to supervise one start-conditioned rollout. To address these limitations, we present CoFL, an end-to-end policy that maps a bird's-eye view (BEV) observation and a language instruction to a continuous flow field for navigation. CoFL reformulates navigation as workspace-conditioned field learning rather than start-conditioned trajectory prediction: it learns local motion vectors at arbitrary BEV locations, turning each scene--instruction annotation into dense spatial control supervision. Trajectories are generated from any start by numerical integration of the predicted field, enabling simple real-time rollout and closed-loop recovery. To enable large-scale training and evaluation, we build a dataset of over 500k BEV image--instruction pairs, each procedurally annotated with a flow field and a trajectory derived from semantic maps built on Matterport3D and ScanNet. Evaluating on strictly unseen scenes, CoFL significantly outperforms modular Vision-Language Model (VLM)-based planners and trajectory generation policies in both navigation precision and safety, while maintaining real-time inference. Finally, we deploy CoFL zero-shot in real-world experiments with BEV observations across multiple layouts, maintaining feasible closed-loop control and a high success rate. |
18 pages, 13 figures |
| Multiple Consistent 2D-3D Mappings for Robust Zero-Shot 3D Visual Grounding | 2026-04-29 | ShowZero-shot 3D Visual Grounding (3DVG) is a critical capability for open-world embodied AI. However, existing methods are fundamentally bottlenecked by the poor quality of open-vocabulary 3D proposals, suffering from inaccurate categories and imprecise geometries, as well as the spatial redundancy of exhaustive multi-view reasoning. To address these challenges, we propose MCM-VG, a novel framework that achieves robust zero-shot 3DVG by explicitly establishing Multiple Consistent 2D-3D Mappings. Instead of passively relying on noisy 3D segments, MCM-VG enforces 2D-3D consistency across three fundamental dimensions to achieve precise target localization and reliable reasoning. First, a Semantic Alignment module corrects category mismatches via LLM-driven query parsing and coarse-to-fine 2D-3D matching. Second, an Instance Rectification module leverages VLM-guided 2D segmentations to reconstruct missing targets, back-projecting these reliable visual priors to establish accurate 3D geometries. Finally, to eliminate spatial redundancy, a Viewpoint Distillation module clusters 3D camera directions to extract optimal frames. By pairing these optimal RGB frames with Bird's Eye View maps into concise visual prompt sets, we formulate the final target disambiguation as a multiple-choice reasoning task for Vision-Language Models. Extensive evaluations on ScanRefer and Nr3D benchmarks demonstrate that MCM-VG sets a new state-of-the-art for zero-shot 3D visual grounding. Remarkably, it achieves 62.0% and 53.6% in Acc@0.25 and Acc@0.5 on ScanRefer, outperforming previous baselines by substantial margins of 6.4% and 4.0%. |
|
| Beyond Shortcuts: Mitigating Visual Illusions in Frozen VLMs via Qualitative Reasoning | 2026-04-29 | ShowWhile Vision-Language Models (VLMs) have achieved state-of-the-art performance in general visual tasks, their perceptual robustness remains remarkably brittle when confronted with optical illusions. These failures are often attributed to shortcut heuristics, where models prioritize linguistic priors and memorized prototypes over direct visual evidence. In this work, we propose Structured Qualitative Inference (SQI), a training-free, data-centric framework designed to fortify visual grounding in frozen VLMs. SQI addresses perceptual anomalies through three systematic modules: (1) Axiomatic Constraint Injection, which suppresses erroneous metric estimations and quantitative hallucinations; (2) Hierarchical Scene Decomposition, which decouples target visual manifolds from complex background distractors; and (3) Counterfactual Self-Verification, an adversarial reasoning step that mitigates confirmation bias. By orchestrating these qualitative constraints at inference time, SQI effectively aligns high-level linguistic reasoning with low-level visual perception. Our framework was evaluated on the DataCV 2026 Challenge (Task I: Classic Illusion Understanding), where it ranked 2nd place overall. Experimental results demonstrate that SQI not only significantly enhances accuracy across diverse illusion categories but also provides superior diagnostic interpretability without any model fine-tuning. Our success underscores the potential of structured qualitative grounding as a robust paradigm for developing next-generation, illusion-resistant vision-language systems. |
4 pag...4 pages, 2 figures, and 1 table. This is a methodology paper for the DataCV 2026 Challenge (CVPR Workshops), Task 1, where our method ranked 2nd |
| CAGE-SGG: Counterfactual Active Graph Evidence for Open-Vocabulary Scene Graph Generation | 2026-04-29 | ShowOpen-vocabulary scene graph generation (SGG) aims to describe visual scenes with flexible and fine-grained relation phrases beyond a fixed predicate vocabulary. While recent vision-language models greatly expand the semantic coverage of SGG, they also introduce a critical reliability issue: predicted relations may be driven by language priors or object co-occurrence rather than grounded visual evidence. In this paper, we propose an evidence-rounded open-vocabulary SGG framework based on counterfactual relation verification. Instead of directly accepting plausible relation proposals, our method verifies whether each candidate relation is supported by relation-pecific visual, geometric, and contextual evidence. Specifically, we first generate open-vocabulary relation candidates with a vision-language proposer, then decompose predicate phrases into soft evidence bases such as support, contact, containment, depth, motion, and state. A relation-conditioned evidence encoder extracts predicate-relevant cues, while a counterfactual verifier tests whether the relation score decreases when necessary vidence is removed and remains stable under irrelevant perturbations. We further introduce contradiction-aware predicate learning and graph-level preference optimization to improve fine-grained discrimination and global graph consistency. Experiments on conventional, open-vocabulary, and panoptic SGG benchmarks show that our method consistently improves standard recall-based metrics, unseen predicate generalization, and counterfactual grounding quality. These results demonstrate that moving from relation generation to relation verification leads to more reliable, interpretable, and evidence-grounded scene graphs. |
This ...This manuscript has been withdrawn by the authors because we found a methodological flaw in the formulation and evaluation of the proposed approach. The issue affects the reliability of the experimental results and the conclusions drawn from them. Therefore, the authors consider the current version unsuitable for citation or further use |
| Beyond Screenshots: Evaluating VLMs' Understanding of UI Animations | 2026-04-28 | ShowAI agents operating on user interfaces must understand how interfaces communicate state and feedback to act reliably. As a core communicative modality, animations are increasingly used in modern interfaces, serving critical functional purposes beyond mere aesthetics. Thus, understanding UI animation is essential for comprehensive interface interpretation. However, recent studies of Vision Language Models (VLMs) for UI understanding have focused primarily on static screenshots, leaving it unclear how well these models handle dynamic UI animations. To address this gap, we created AniMINT, a novel dataset of 300 densely annotated UI animation videos. We systematically evaluate state-of-the-art VLMs on UI animation understanding, including their abilities to perceive the animation effects, identify animation purposes, and interpret animation meaning. Our results show that VLMs can reliably detect primitive motion. However, their high-level animation interpretation remains inconsistent, with substantial gaps relative to human performance. Finally, we use Motion, Context, and Perceptual Cues (MCPC) to probe factors affecting VLM performance, revealing key bottlenecks and directions for future improvement. |
Accep...Accepted at ACL 2026 Findings |
| Time Blindness: Why Video-Language Models Can't See What Humans Can? | 2026-04-28 | ShowRecent advances in vision-language models (VLMs) have made impressive strides in understanding spatio-temporal relationships in videos. However, when spatial information is obscured, these models struggle to capture purely temporal patterns. We introduce |
Accep...Accepted at IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 Project page at https://timeblindness.github.io |
| Robot Planning and Situation Handling with Active Perception | 2026-04-28 | ShowCurrent robots are capable of computing plans to accomplish complex tasks. However, real-world environments are inherently open and dynamic, and unforeseen situations frequently arise during plan execution, such as jamming doors and fallen objects on the floor. These situations may result from the robot's own action failures or from external disturbances, such as human activities. Detecting and handling such execution - time situations remains a significant challenge, limiting those robots' ability to achieve long-term autonomy. In this paper, we develop a planning and situation-handling framework, called VAP-TAMP, that enables robots to actively perceive and address unforeseen situations during plan execution. VAP-TAMP leverages action knowledge to strategically prompt vision-language models for active view selection and situation assessment, while constructing and reasoning over scene graphs for integrated task and motion planning. We evaluated VAP-TAMP using service tasks in simulation and on a mobile manipulation platform. |
|
| VIGNETTE: Socially Grounded Bias Evaluation for Vision-Language Models | 2026-04-28 | ShowWhile bias in large language models (LLMs) is well-studied, similar concerns in vision-language models (VLMs) have received comparatively less attention. Existing VLM bias studies often focus on portrait-style images and gender-occupation associations, overlooking broader and more complex social stereotypes and their implied harm. This work introduces VIGNETTE, a large-scale VQA benchmark with 30M+ images for evaluating bias in VLMs through a question-answering framework spanning four directions: factuality, perception, stereotyping, and decision making. Beyond narrowly-centered studies, we assess how VLMs interpret identities in contextualized settings, revealing how models make trait and capability assumptions and exhibit patterns of discrimination. Drawing from social psychology, we examine how VLMs connect visual identity cues to trait and role-based inferences, encoding social hierarchies, through biased selections. Our findings uncover subtle, multifaceted, and surprising stereotypical patterns, offering insights into how VLMs construct social meaning from inputs. |
Accepted to ACL 2026 |
| QCalEval: Benchmarking Vision-Language Models for Quantum Calibration Plot Understanding | 2026-04-28 | ShowQuantum computing calibration depends on interpreting experimental data, and calibration plots provide the most universal human-readable representation for this task, yet no systematic evaluation exists of how well vision-language models (VLMs) interpret them. We introduce QCalEval, the first VLM benchmark for quantum calibration plots: 243 samples across 87 scenario types from 22 experiment families, spanning superconducting qubits and neutral atoms, evaluated on six question types in both zero-shot and in-context learning settings. The best general-purpose zero-shot model reaches a mean score of 72.3, and many open-weight models degrade under multi-image in-context learning, whereas frontier closed models improve substantially. A supervised fine-tuning ablation at the 9-billion-parameter scale shows that SFT improves zero-shot performance but cannot close the multimodal in-context learning gap. As a reference case study, we release NVIDIA Ising Calibration 1, an open-weight model based on Qwen3.5-35B-A3B that reaches 74.7 zero-shot average score. |
|
| Personalization Toolkit: Training Free Personalization of Large Vision Language Models | 2026-04-28 | ShowPersonalization of Large Vision-Language Models (LVLMs) involves customizing models to recognize specific users or object instances and to generate contextually tailored responses. Existing approaches rely on time-consuming training for each item, making them impractical for real-world deployment, as reflected in current personalization benchmarks limited to object-centric single-concept evaluations. In this paper, we present a novel training-free approach to LVLM personalization called \ours. We introduce a comprehensive, real-world benchmark designed to rigorously evaluate various aspects of the personalization task. \ours leverages pre-trained vision foundation models to extract distinctive features, applies retrieval-augmented generation (RAG) techniques to identify instances within visual inputs, and employs visual prompting strategies to guide model outputs. Our model-agnostic vision toolkit enables efficient and flexible multi-concept personalization across both images and videos, without any additional training. We achieve state-of-the-art results, surpassing existing training-based methods. |
Accep...Accepted at Transactions on Machine Learning Research (TMLR) 2026 |
| Instruction-Evidence Contrastive Dual-Stream Decoding for Grounded Vision-Language Reasoning | 2026-04-28 | ShowVision-Language Models (VLMs) exhibit strong performance in instruction following and open-ended vision-language reasoning, yet they frequently generate fluent outputs that are weakly grounded in visual evidence. Prior works have shown that instruction prompting further worsens this issue by amplifying language priors, especially when the visual signal is uncertain or ambiguous. To address this challenge, we propose a decoding framework that explicitly balances linguistic informativeness and visual faithfulness during generation. Our method, Instruction-Evidence Contrastive Dual-Stream Decoding (IECD2), maintains two parallel probability distributions of tokens at each decoding step: an instruction-driven stream that promotes expressive and informative responses, and an evidence-driven stream that enforces strict grounding in the image. These two streams are adaptively fused using a symmetric KL-based contrast-based gate, which suppresses tokens favored by language priors but unsupported by visual evidence, while preserving them when both distributions agree. We evaluate IECD2 on multiple datasets spanning various generative vision-language reasoning tasks such as captioning and visual question answering, including POPE, MME, VQAv2, AMBER, MS-COCO, and LLaVA-Bench. IECD2 demonstrates consistent improvements in task accuracy and reasoning performance, alongside a substantial reduction in hallucination across all evaluation metrics compared to state-of-the-art decoding approaches. |
|
| CodeOCR: On the Effectiveness of Vision Language Models in Code Understanding | 2026-04-28 | ShowLarge Language Models (LLMs) have achieved remarkable success in source code understanding, yet as software systems grow in scale, computational efficiency has become a critical bottleneck. Currently, these models rely on a text-based paradigm that treats source code as a linear sequence of tokens, which leads to a linear increase in context length and associated computational costs. The rapid advancement of Multimodal LLMs (MLLMs) introduces an opportunity to optimize efficiency by representing source code as rendered images. Unlike text, which is difficult to compress without losing semantic meaning, the image modality is inherently suitable for compression. By adjusting resolution, images can be scaled to a fraction of their original token cost while remaining recognizable to vision-capable models. To explore the feasibility of this approach, we conduct the first systematic study on the effectiveness of MLLMs for code understanding. Our experiments reveal that: (1) MLLMs can effectively understand code with substantial token reduction, achieving up to 8x compression; (2) MLLMs can effectively leverage visual cues such as syntax highlighting, improving code completion performance under 4x compression; and (3) Code-understanding tasks like clone detection exhibit exceptional resilience to visual compression, with some compression ratios even slightly outperforming raw text inputs. Our findings highlight both the potential and current limitations of MLLMs in code understanding, which points out a shift toward image-modality code representation as a pathway to more efficient inference. |
Accep...Accepted to ISSTA 2026. Code and data are available at https://github.com/YerbaPage/CodeOCR |
| MemeScouts@LT-EDI 2026: Asking the Right Questions -- Prompted Weak Supervision for Meme Hate Speech Detection | 2026-04-28 | ShowDetecting hate speech in memes is challenging due to their multimodal nature and subtle, culturally grounded cues such as sarcasm and context. While recent vision-language models (VLMs) enable joint reasoning over text and images, end-to-end prompting can be brittle, as a single prediction must resolve target, stance, implicitness, and irony. These challenges are amplified in multilingual settings. We propose a prompted weak supervision (PWS) approach that decomposes meme understanding into targeted, question-based labeling functions with constrained answer options for homophobia and transphobia detection in the LT-EDI 2026 shared task. Using a quantized Qwen3-VLM to extract features by answering targeted questions, our method outperforms direct VLM classification, with substantial gains for Chinese and Hindi, ranking 1st in English, 2nd in Chinese, and 3rd in Hindi. Iterative refinement via error-driven LF expansion and feature pruning reduces redundancy and improves generalization. Our results highlight the effectiveness of prompted weak supervision for multilingual multimodal hate speech detection. |
Accep...Accepted at Sixth Workshop on Language Technology for Equality, Diversity and Inclusion at ACL2026 (LT-EDI@ACL26) |
| A Comparative Study in Surgical AI: Datasets, Foundation Models, and Barriers to Med-AGI | 2026-04-28 | ShowRecent Artificial Intelligence (AI) models have matched or exceeded human experts in several benchmarks of biomedical task performance, but surgical benchmarks in particular are often missing from prominent medical benchmark suites (specifically, those requiring visual recognition). Since surgery requires integrating disparate tasks, generally-capable AI models could be particularly attractive as a collaborative tool if performance could be improved. On the one hand, the canonical approach of scaling architecture size and training data is attractive, especially since there are millions of hours of surgical video data generated per year. On the other hand, preparing surgical data for AI training requires significantly higher levels of professional expertise, and training on that data requires expensive computational resources. These trade-offs paint an uncertain picture of whether and to-what-extent modern AI could aid surgical practice. In this paper, we explore this question through a case study of surgical tool detection using state-of-the-art AI methods available in 2026. We demonstrate that even with multi-billion parameter models and extensive training, current Vision Language Models fall short in the seemingly simple task of tool detection in neurosurgery. Additionally, we show scaling experiments indicating that increasing model size and training time only leads to diminishing improvements in relevant performance metrics. Thus, our experiments suggest that current models could still face significant obstacles in surgical use cases. Moreover, some obstacles cannot be simply ``scaled away'' with additional compute and persist across diverse model architectures, raising the question of whether data and label availability are the only limiting factors. We discuss the main contributors to these constraints and advance potential solutions. |
|
| Prefill-Time Intervention for Mitigating Hallucination in Large Vision-Language Models | 2026-04-28 | ShowLarge Vision-Language Models (LVLMs) have achieved remarkable progress in visual-textual understanding, yet their reliability is critically undermined by hallucinations, i.e., the generation of factually incorrect or inconsistent responses. While recent studies using steering vectors demonstrated promise in reducing hallucinations, a notable challenge remains: they inadvertently amplify the severity of residual hallucinations. We attribute this to their exclusive focus on the decoding stage, where errors accumulate autoregressively and progressively worsen subsequent hallucinatory outputs. To address this, we propose Prefill-Time Intervention (PTI), a novel steering paradigm that intervenes only once during the prefill stage, enhancing the initial Key-Value (KV) cache before error accumulation occurs. Specifically, PTI is modality-aware, deriving distinct directions for visual and textual representations. This intervention is decoupled to steer keys toward visually-grounded objects and values to filter background noise, correcting hallucination-prone representations at their source. Extensive experiments demonstrate PTI's significant performance in mitigating hallucinations and its generalizability across diverse decoding strategies, LVLMs, and benchmarks. Moreover, PTI is orthogonal to existing decoding-stage methods, enabling plug-and-play integration and further boosting performance. Code is available at: https://github.com/huaiyi66/PTI. |
Accep...Accepted by CVPR 2026 |
| PromptEvolver: Prompt Inversion through Evolutionary Optimization in Natural-Language Space | 2026-04-28 | ShowText-to-image generation has progressed rapidly, but faithfully generating complex scenes requires extensive trial-and-error to find the exact prompt. In the prompt inversion task, the goal is to recover a textual prompt that can faithfully reconstruct a given target image. Currently, existing methods frequently yield suboptimal reconstructions and produce unnatural, hard-to-interpret prompts that hinder transparency and controllability. In this work, we present PromptEvolver, a prompt inversion approach that generates natural-language prompts while achieving high-fidelity reconstructions of the target image. Our method uses a genetic algorithm to optimize the prompt, leveraging a strong vision-language model to guide the evolution process. Importantly, it works on black-box generation models by requiring only image outputs. Finally, we evaluate PromptEvolver across multiple prompt inversion benchmarks and show that it consistently outperforms competing methods. |
|
| SnapGuard: Lightweight Prompt Injection Detection for Screenshot-Based Web Agents | 2026-04-28 | ShowWeb agents have emerged as an effective paradigm for automating interactions with complex web environments, yet remain vulnerable to prompt injection attacks that embed malicious instructions into webpage content to induce unintended actions. This threat is further amplified for screenshot-based web agents, which operate on rendered visual webpages rather than structured textual representations, making predominant text-centric defenses ineffective. Although multimodal detection methods have been explored, they often rely on large vision-language models (VLMs), incurring significant computational overhead. The bottleneck lies in the complexity of modern webpages: VLMs must comprehend the global semantics of an entire page, resulting in substantial inference time and GPU memory usage. This raises a critical question: can we detect prompt injection attacks from screenshots in a lightweight manner? In this paper, we observe that injected webpages exhibit distinct characteristics compared to benign ones from both visual and textual perspectives. Building on this insight, we propose SnapGuard, a lightweight yet accurate method that reformulates prompt injection detection as multimodal representation analysis over webpage screenshots. SnapGuard leverages two complementary signals: a visual stability indicator that identifies abnormally smooth gradient distributions induced by malicious content, and action-oriented textual signals recovered via contrast-polarity reversal. Extensive evaluations across eight attacks and two benign settings demonstrate that SnapGuard achieves an F1 score of 0.75, outperforming GPT-4o-prompt while being 8x faster (1.81s vs. 14.50s) and introducing no additional memory overhead. |
10 pages, 7 figures |
| HuM-Eval: A Coarse-to-Fine Framework for Human-Centric Video Evaluation | 2026-04-28 | ShowVideo generation models have developed rapidly in recent years, where generating natural human motion plays a pivotal role. However, accurately evaluating the quality of generated human motion video remains a significant challenge. Existing evaluation metrics primarily focus on global scene statistics, often overlooking fine-grained human details and consequently failing to align with human subjective preference. To bridge this gap, we propose HuM-Eval, a novel human-centric evaluation framework that adopts a coarse-to-fine strategy. Specifically, our framework first utilizes a Vision Language Model to perform a coarse assessment of global video quality. It then proceeds to a fine-grained analysis, using 2D pose to verify anatomical correctness and 3D human motion to evaluate motion stability. Extensive experiments demonstrate that HuM-Eval achieves an average human correlation of 58.2%, outperforming state-of-the-art baselines. Furthermore, we introduce HuM-Bench, a comprehensive benchmark comprising 1,000 diverse prompts, and conduct a detailed evaluation of existing text-to-video models, paving the way for next-generation human motion generation. |
Accep...Accepted to the 2026 IEEE International Conference on Multimedia and Expo (ICME 2026) |
| Zoom In, Reason Out: Efficient Far-field Anomaly Detection in Expressway Surveillance Videos via Focused VLM Reasoning Guided by Bayesian Inference | 2026-04-28 | ShowExpressway video anomaly detection is essential for safety management. However, identifying anomalies across diverse scenes remains challenging, particularly for far-field targets exhibiting subtle abnormal vehicle motions. While Vision-Language Models (VLMs) demonstrate strong semantic reasoning capabilities, processing global frames causes attention dilution for these far-field objects and incurs prohibitive computational costs. To address these issues, we propose VIBES, an asynchronous collaborative framework utilizing VLMs guided by Bayesian inference. Specifically, to overcome poor generalization across varying expressway environments, we introduce an online Bayesian inference module. This module continuously evaluates vehicle trajectories to dynamically update the probabilistic boundaries of normal driving behaviors, serving as an asynchronous trigger to precisely localize anomalies in space and time. Instead of processing the continuous video stream, the VLM processes only the localized visual regions indicated by the trigger. This targeted visual input prevents attention dilution and enables accurate semantic reasoning. Extensive evaluations demonstrate that VIBES improves detection accuracy for far-field anomalies and reduces computational overhead, achieving high real-time efficiency and explainability while demonstrating generalization across diverse expressway conditions. |
|
| Images Amplify Misinformation Sharing in Vision-Language Models | 2026-04-28 | ShowAs language and vision-language models (VLMs) become central to information access and online interaction, concerns grow about their potential to amplify misinformation. Human studies show that images boost the perceived credibility and shareability of information, raising the question of whether VLMs exhibit the same vulnerability. We present the first study examining how images influence VLMs' propensity to reshare news content, how this effect varies across model families, and how persona conditioning and content attributes modulate such behavior. We develop a jailbreaking-inspired prompting strategy that bypasses VLMs' default refusals to engage with controversial news, allowing them to generate resharing decisions across diverse topics and elicited traits, including antisocial ones. We evaluate four state-of-the-art VLMs on a novel multimodal dataset of fact-checked political news from PolitiFact, paired with images and ground-truth veracity labels. Our experiments show that image presence increases resharing rates by 14.5% for false news and 5.3% for true news. Persona conditioning further modulates this effect: Dark Triad traits amplify resharing of false news, whereas Republican-aligned profiles reduce sensitivity to veracity. Among the tested models, Claude-3-Haiku demonstrates the greatest robustness to visual misinformation. These findings reveal that VLMs replicate human-like biases in response to images, underscoring emerging risks for multimodal AI systems. They point to the need for evaluation frameworks and mitigation strategies that account for visual influence and persona-driven variability, particularly in sociotechnical settings where AI systems shape public discourse and information sharing. |
Accep...Accepted for oral presentation at ICWSM 2026 |
| DRAGON: A Benchmark for Evidence-Grounded Visual Reasoning over Diagrams | 2026-04-28 | ShowDiagram question answering (DQA) requires models to interpret structured visual representations such as charts, maps, infographics, circuit schematics, and scientific diagrams. Recent vision-language models (VLMs) often achieve high answer accuracy on these tasks, yet correct answers do not guarantee that models ground their reasoning in the diagram regions that support the prediction. Models may instead rely on textual correlations or dataset artifacts without identifying the visual evidence required to verify the answer. This limitation prevents reliable evaluation of diagram reasoning and reduces interpretability. We introduce DRAGON, a benchmark for evaluating evidence-grounded visual reasoning in diagrams. Given a diagram, a question, and the correct answer, a model must predict bounding boxes that correspond to the visual elements required to justify the answer. These evidence regions may include answer-bearing components, textual labels, legends, axes, connectors, and other supporting structures involved in the reasoning process. The DRAGON dataset contains 11,664 annotated question instances collected from six diagram QA datasets: ChartQA, Circuit-VQA, InfographicsVQA, MapIQ, MapWise, and AI2D. We release a 2,445-instance benchmark test set with human-verified reasoning evidence annotations and a standardized evaluation framework. We evaluate eight recent VLMs and analyze their ability to localize reasoning evidence across diverse diagram domains. DRAGON enables systematic evaluation of diagram reasoning and supports future research on models that ground their predictions in visual evidence. |
22 Pages, 14 Figures |
| SketchVLM: Vision language models can annotate images to explain thoughts and guide users | 2026-04-28 | ShowWhen answering questions about images, humans naturally point, label, and draw to explain their reasoning. In contrast, modern vision-language models (VLMs) such as Gemini-3-Pro and GPT-5 only respond with text, which can be difficult for users to verify. We present SketchVLM, a training-free, model-agnostic framework that enables VLMs to produce non-destructive, editable SVG overlays on the input image to visually explain their answers. Across seven benchmarks spanning visual reasoning (maze navigation, ball-drop trajectory prediction, and object counting) and drawing (part labeling, connecting-the-dots, and drawing shapes around objects), SketchVLM improves visual reasoning task accuracy by up to +28.5 percentage points and annotation quality by up to 1.48x relative to image-editing and fine-tuned sketching baselines, while also producing annotations that are more faithful to the model's stated answer. We find that single-turn generation already achieves strong accuracy and annotation quality, and multi-turn generation opens up further opportunities for human-AI collaboration. An interactive demo and code are at https://sketchvlm.github.io/. |
|
| From Scene to Object: Text-Guided Dual-Gaze Prediction | 2026-04-28 | ShowInterpretable driver attention prediction is crucial for human-like autonomous driving. However, existing datasets provide only scene-level global gaze rather than fine-grained object-level annotations, inherently failing to support text-grounded cognitive modeling. Consequently, while Vision-Language Models (VLMs) hold great potential for semantic reasoning, this critical data limitations leads to severe text-vision decoupling and visual-bias hallucinations. To break this bottleneck and achieve precise object-level attention prediction, this paper proposes a novel dual-branch gaze prediction framework, establishing a complete paradigm from data construction to model architecture. First, we construct G-W3DA, a object-level driver attention dataset. By integrating a multimodal large language model with the Segment Anything Model 3 (SAM3), we decouple macroscopic heatmaps into object-level masks under rigorous cross-validation, fundamentally eliminating annotation hallucinations. Building upon this high-quality data foundation, we propose the DualGaze-VLM architecture. This architecture extracts the hidden states of semantic queries and dynamically modulates visual features via a Condition-Aware SE-Gate, achieving intent-driven precise spatial anchoring. Extensive experiments on the W3DA benchmark demonstrate that DualGaze-VLM consistently surpasses existing state-of-the-art (SOTA) models in spatial alignment metrics, notably achieving up to a 17.8% improvement in Similarity (SIM) under safety-critical scenarios. Furthermore, a visual Turing test reveals that the attention heatmaps generated by DualGaze-VLM are perceived as authentic by 88.22% of human evaluators, proving its capability to generate rational cognitive priors. |
|
| SARE: Sample-wise Adaptive Reasoning for Training-free Fine-grained Visual Recognition | 2026-04-28 | ShowRecent advances in Large Vision-Language Models (LVLMs) have enabled training-free Fine-Grained Visual Recognition (FGVR). However, effectively exploiting LVLMs for FGVR remains challenging due to the inherent visual ambiguity of subordinate-level categories. Existing methods predominantly adopt either retrieval-oriented or reasoning-oriented paradigms to tackle this challenge, but both are constrained by two fundamental limitations:(1) They apply the same inference pipeline to all samples without accounting for uneven recognition difficulty, thereby leading to suboptimal accuracy and efficiency; (2) The lack of mechanisms to consolidate and reuse error-specific experience causes repeated failures on similar challenging cases. To address these limitations, we propose SARE, a Sample-wise Adaptive textbfREasoning framework for training-free FGVR. Specifically, SARE adopts a cascaded design that combines fast candidate retrieval with fine-grained reasoning, invoking the latter only when necessary. In the reasoning process, SARE incorporates a self-reflective experience mechanism that leverages past failures to provide transferable discriminative guidance during inference, without any parameter updates. Extensive experiments across 14 datasets substantiate that SARE achieves state-of-the-art performance while substantially reducing computational overhead. |
prepr...preprint, under review |
| MPR-GUI: Benchmarking and Enhancing Multilingual Perception and Reasoning in GUI Agents | 2026-04-28 | ShowLarge Vision-Language Models (LVLMs) have shown strong potential as multilingual Graphical User Interface (GUI) agents, as evidenced by existing GUI benchmarks. However, these benchmarks exhibit two primary limitations: (1) although Perception and Reasoning (P&R) capabilities are fundamental for GUI agents, current benchmarks lack fine-grained diagnostics to identify which specific capabilities lead to task failures, hindering targeted improvements; (2) existing benchmarks fail to provide a strictly aligned cross-lingual evaluation environment, introducing confounding factors that prevent isolating the language impact on GUI agent performance. To address these issues, we propose the Multilingual P&R GUI Benchmark (MPR-GUI-Bench), featuring strictly aligned environments across six languages and eight fine-grained P&R tasks. Our benchmark reveals consistent P&R gaps between English and non-English settings, particularly on reasoning-intensive tasks. To leverage the superior English P&R capabilities for bridging cross-lingual gaps, we identify layers sensitive to language and propose GUI-XLI, a GUI Cross-Lingual Intervention method that aligns non-English hidden states with their English counterparts at these layers during inference. Experiments show that GUI-XLI effectively reduces the cross-lingual gaps, with an average gain of 6.5% in non-English settings. |
35pages, 15figures |
| Latent Anomaly Knowledge Excavation: Unveiling Sparse Sensitive Neurons in Vision-Language Models | 2026-04-28 | ShowLarge-scale vision-language models (VLMs) exhibit remarkable zero-shot capabilities, yet the internal mechanisms driving their anomaly detection (AD) performance remain poorly understood. Current methods predominantly treat VLMs as black-box feature extractors, assuming that anomaly-specific knowledge must be acquired through external adapters or memory banks. In this paper, we challenge this assumption by arguing that anomaly knowledge is intrinsically embedded within pre-trained models but remains latent and under-activated. We hypothesize that this knowledge is concentrated within a sparse subset of anomaly-sensitive neurons. To validate this, we propose latent anomaly knowledge excavation (LAKE), a training-free framework that identifies and elicits these critical neuronal signals using only a minimal set of normal samples. By isolating these sensitive neurons, LAKE constructs a highly compact normality representation that integrates visual structural deviations with cross-modal semantic activations. Extensive experiments on industrial AD benchmarks demonstrate that LAKE achieves state-of-the-art performance while providing intrinsic, neuron-level interpretability. Ultimately, our work advocates for a paradigm shift: redefining anomaly detection as the targeted activation of latent pre-trained knowledge rather than the acquisition of a downstream task. |
|
| DIAL: Decoupling Intent and Action via Latent World Modeling for End-to-End VLA | 2026-04-28 | ShowThe development of Vision-Language-Action (VLA) models has been significantly accelerated by pre-trained Vision-Language Models (VLMs). However, most existing end-to-end VLAs treat the VLM primarily as a multimodal encoder, directly mapping vision-language features to low-level actions. This paradigm underutilizes the VLM's potential in high-level decision making and introduces training instability, frequently degrading its rich semantic representations. To address these limitations, we introduce DIAL, a framework bridging high-level decision making and low-level motor execution through a differentiable latent intent bottleneck. Specifically, a VLM-based System-2 performs latent world modeling by synthesizing latent visual foresight within the VLM's native feature space; this foresight explicitly encodes intent and serves as the structural bottleneck. A lightweight System-1 policy then decodes this predicted intent together with the current observation into precise robot actions via latent inverse dynamics. To ensure optimization stability, we employ a two-stage training paradigm: a decoupled warmup phase where System-2 learns to predict latent futures while System-1 learns motor control under ground-truth future guidance within a unified feature space, followed by seamless end-to-end joint optimization. This enables action-aware gradients to refine the VLM backbone in a controlled manner, preserving pre-trained knowledge. Extensive experiments on the RoboCasa GR1 Tabletop benchmark show that DIAL establishes a new state-of-the-art, achieving superior performance with 10x fewer demonstrations than prior methods. Furthermore, by leveraging heterogeneous human demonstrations, DIAL learns physically grounded manipulation priors and exhibits robust zero-shot generalization to unseen objects and novel configurations during real-world deployment on a humanoid robot. |
Proje...Project page: https://xpeng-robotics.github.io/dial |
| Genie Sim 3.0 : A High-Fidelity Comprehensive Simulation Platform for Humanoid Robot | 2026-04-28 | ShowThe development of robust and generalizable robot learning models is critically contingent upon the availability of large-scale, diverse training data and reliable evaluation benchmarks. Collecting data in the physical world poses prohibitive costs and scalability challenges, and prevailing simulation benchmarks frequently suffer from fragmentation, narrow scope, or insufficient fidelity to enable effective sim-to-real transfer. To address these challenges, we introduce Genie Sim 3.0, a unified simulation platform for robotic manipulation. We present Genie Sim Generator, a large language model (LLM)-powered tool that constructs high-fidelity scenes from natural language instructions. Its principal strength resides in rapid and multi-dimensional generalization, facilitating the synthesis of diverse environments to support scalable data collection and robust policy evaluation. We introduce the first benchmark that pioneers the application of LLM for automated evaluation. It leverages LLM to mass-generate evaluation scenarios and employs Vision-Language Model (VLM) to establish an automated assessment pipeline. We also release an open-source dataset comprising more than 10,000 hours of synthetic data across over 200 tasks. Through systematic experimentation, we validate the robust zero-shot sim-to-real transfer capability of our open-source dataset, demonstrating that synthetic data can server as an effective substitute for real-world data under controlled conditions for scalable policy training. For code and dataset details, please refer to: https://github.com/AgibotTech/genie_sim. |
|
| One Perturbation, Two Failure Modes: Probing VLM Safety via Embedding-Guided Typographic Perturbations | 2026-04-28 | ShowTypographic prompt injection exploits vision language models' (VLMs) ability to read text rendered in images, posing a growing threat as VLMs power autonomous agents. Prior work typically focus on maximizing attack success rate (ASR) but does not explain \emph{why} certain renderings bypass safety alignment. We make two contributions. First, an empirical study across four VLMs including GPT-4o and Claude, twelve font sizes, and ten transformations reveals that multimodal embedding distance strongly predicts ASR ( |
|
| Dynamic Decision Learning: Test-Time Evolution for Abnormality Grounding in Rare Diseases | 2026-04-27 | ShowClinical abnormality grounding for rare diseases is often hindered by data scarcity, making supervised fine-tuning impractical and single-pass inference highly unstable. We propose Dynamic Decision Learning (DDL), a framework that enables frozen large vision-language models (LVLMs) to refine their decisions across both language and visual spaces by optimizing instructions and consolidating predictions under visual perturbations. This process improves localization quality and produces a consensus-based reliability score that quantifies model confidence. Results on brain imaging benchmarks, including a rare-disease dataset with 281 pathology types across models ranging from 3B to 72B parameters, show that DDL improves mAP@75 by up to 105% on rare-disease cases and outperforms adaptation baselines and supervised fine-tuning. Furthermore, DDL demonstrates stronger calibration between reliability scores and localization accuracy under severe distribution shifts and increasing task difficulty. Code is available at: https://lijunrio.github.io/DDL/ |
|
| Agentic AI for Remote Sensing: Technical Challenges and Research Directions | 2026-04-27 | ShowEarth Observation (EO) is moving beyond static prediction toward multi-step analytical workflows that require coordinated reasoning over data, tools, and geospatial state. While foundation models and vision-language models have expanded representation learning and language-grounded interaction for remote sensing, and agentic AI has demonstrated long-horizon reasoning and external tool use, EO is not a straightforward extension of generic agentic AI. EO workflows operate over georeferenced, multi-modal, and temporally structured data, where operations such as reprojection, resampling, compositing, and aggregation actively transform the underlying state and can constrain subsequent analysis. As a result, errors may propagate silently across steps, and correctness depends not only on internal coherence, but also on geospatial consistency, temporally valid comparisons, and physical validity. This position paper argues that these challenges are structural rather than incidental. We identify the implicit assumptions commonly made in generic agentic models, analyze how they break in geospatial workflows, and characterize the resulting failure modes in multi-step EO pipelines. We then outline design principles for EO-native agents centered on structured geospatial state, tool-aware reasoning, verifier-guided execution, and learning objectives aligned with geospatial and physical validity. Finally, we present research directions spanning EO-specific benchmarks, hybrid supervised and reinforcement learning, constrained self-improvement, and trajectory-level evaluation beyond final-answer accuracy. Building reliable geospatial agents therefore requires rethinking agent design around the physical, geospatial, and workflow constraints that govern EO analysis. |
31 pa...31 pages. Position Paper |
| Interactive Episodic Memory with User Feedback | 2026-04-27 | ShowIn episodic memory with natural language queries (EM-NLQ), a user may ask a question (e.g., "Where did I place the mug?") that requires searching a long egocentric video, captured from the user's perspective, to find the moment that answers it. However, queries can be ambiguous or incomplete, leading to incorrect responses. Current methods ignore this key aspect and address EM-NLQ in a one-shot setup, limiting their applicability in real-world scenarios. In this work, we address this gap and introduce the Episodic Memory with Questions and Feedback task (EM-QnF). Here, the user can provide feedback on the model's initial prediction or add more information (e.g., "Before this. I'm looking for the big blue mug not the white one"), helping the model refine its predictions interactively. To this end, we collect datasets for feedback-based interaction and propose a lightweight training scheme that avoids expensive sequential optimization. We also introduce a plug-and-play Feedback ALignment Module (FALM) that enables existing EM-NLQ models to incorporate user feedback effectively. Our approach significantly improves over the state of the art on three challenging benchmarks and is better than or competitive with commercial large vision-language models while remaining efficient. Evaluation with human-generated feedback shows that it generalizes well to real-world scenarios. |
Accep...Accepted to CVPR 2026. Project Page: https://nsubedi11.github.io/refocus |
| VERTIGO: Visual Preference Optimization for Cinematic Camera Trajectory Generation | 2026-04-27 | ShowCinematic camera control relies on a tight feedback loop between director and cinematographer, where camera motion and framing are continuously reviewed and refined. Recent generative camera systems can produce diverse, text-conditioned trajectories, but they lack this "director in the loop" and have no explicit supervision of whether a shot is visually desirable. This results in in-distribution camera motion but poor framing, off-screen characters, and undesirable visual aesthetics. In this paper, we introduce VERTIGO, the first framework for visual preference optimization of camera trajectory generators. Our framework leverages a real-time graphics engine (Unity) to render 2D visual previews from generated camera motion. A cinematically fine-tuned vision-language model then scores these previews using our proposed cyclic semantic similarity mechanism, which aligns renders with text prompts. This process provides the visual preference signals for Direct Preference Optimization (DPO) post-training. Both quantitative evaluations and user studies on Unity renders and diffusion-based Camera-to-Video pipelines show consistent gains in condition adherence, framing quality, and perceptual realism. Notably, VERTIGO reduces the character off-screen rate from 38% to nearly 0% while preserving the geometric fidelity of camera motion. User study participants further prefer VERTIGO over baselines across composition, consistency, prompt adherence, and aesthetic quality, confirming the perceptual benefits of our visual preference post-training. |
28 pages, 10 figures |
| World-R1: Reinforcing 3D Constraints for Text-to-Video Generation | 2026-04-27 | ShowRecent video foundation models demonstrate impressive visual synthesis but frequently suffer from geometric inconsistencies. While existing methods attempt to inject 3D priors via architectural modifications, they often incur high computational costs and limit scalability. We propose World-R1, a framework that aligns video generation with 3D constraints through reinforcement learning. To facilitate this alignment, we introduce a specialized pure text dataset tailored for world simulation. Utilizing Flow-GRPO, we optimize the model using feedback from pre-trained 3D foundation models and vision-language models to enforce structural coherence without altering the underlying architecture. We further employ a periodic decoupled training strategy to balance rigid geometric consistency with dynamic scene fluidity. Extensive evaluations reveal that our approach significantly enhances 3D consistency while preserving the original visual quality of the foundation model, effectively bridging the gap between video generation and scalable world simulation. |
Proje...Project Page: https://aka.ms/world-r1, Code: https://github.com/microsoft/World-R1 |
| SPEAR-1: Scaling Beyond Robot Demonstrations via 3D Understanding | 2026-04-27 | ShowRobotic Foundation Models (RFMs) hold great promise as generalist, end-to-end systems for robot control. Yet their ability to generalize across new environments, tasks, and embodiments remains limited. We argue that a major bottleneck lies in their foundations: most RFMs are built by fine-tuning internet-pretrained Vision-Language Models (VLMs). However, these VLMs are trained on 2D image-language tasks and lack the 3D spatial reasoning inherently required for embodied control in the 3D world. Bridging this gap directly with large-scale robotic data is costly and difficult to scale. Instead, we propose to enrich easy-to-collect non-robotic image data with 3D annotations and enhance a pretrained VLM with 3D understanding capabilities. Following this strategy, we train SPEAR-VLM, a 3D-aware VLM that infers object coordinates in 3D space from a single 2D image. Building on SPEAR-VLM, we introduce our main contribution, |
|
| Self-Rewarding Vision-Language Model via Reasoning Decomposition | 2026-04-27 | ShowVision-Language Models (VLMs) often suffer from visual hallucinations: generating things that are not consistent with visual inputs and language shortcuts, where they skip the visual part and just rely on text priors. These issues arise because most post training methods for VLMs rely on simple verifiable answer matching and supervise only final outputs, leaving intermediate visual reasoning without explicit guidance. As a result, VLMs receive sparse visual signals and often learn to prioritize language based reasoning over visual perception. We introduce Vision SR1, a three stage self rewarding reinforcement learning method that improves visual reasoning without relying on external visual supervision. Vision SR1 decomposes VLM reasoning into two components: visual reasoning and language reasoning, where the model is first prompted to produce self-contained visual descriptions sufficient to answer the question without referring back to the input image, before jointly optimizing both visual and language reasoning through our multi reward loss objective. To validate this self containment, the same VLM model is reprompted to perform language reasoning using only the generated visual reasoning as input to compute visual reward. The final reward is computed through a decoupled reward-advantage framework, where visual reward and language reasoning reward each have their advantages calculated separately. Our experiments show that Vision SR1 improves visual reasoning, mitigates visual hallucinations, and reduces reliance on language shortcuts across diverse vision language tasks, while being more efficient than methods that rely on external visual reward models, which require additional GPUs to host. In contrast, Vision SR1 introduces no extra GPU overhead beyond that of standard training. |
16 pa...16 pages, two figures |