Bg Remover

Bg Remover — AI-powered image generation, transformation, and creative exploration.

---

Topics: image-processing · background-removal · computer-vision · digital-image · dsp-techniques · image-denoising · image-segmentation · object-detection · yolo

Overview

Bg Remover is a generative image application that leverages diffusion models or GAN-based pipelines to create, transform, or stylise images from natural language prompts and visual inputs. It provides an interactive interface for experimenting with generative AI's visual capabilities.

The application supports text-to-image generation with configurable parameters (guidance scale, steps, seed, size), image-to-image transformation with adjustable strength, and a gallery for browsing and comparing generated outputs. All generation parameters are exposed in the UI for maximum experimental flexibility.

The backend can be configured to use cloud APIs (fal.ai, Replicate, Stability AI) for instant availability without local GPU, or local Diffusers pipelines for offline operation and maximum customisation.

---

Motivation

Generative image models have fundamentally changed what's possible in visual creativity. This project was built to make these capabilities accessible and explorable without requiring deep ML expertise or expensive GPU infrastructure.

---

Architecture

User Prompt + Parameters
        │
  Diffusion Pipeline (API or local)
        │
  Generated Image + Metadata
        │
  Gallery Display + Download

---

Features

Text-to-Image Generation

Generate high-quality images from natural language prompts with full parameter control over guidance, steps, seed, and dimensions.

Negative Prompt Support

Specify elements to exclude from generated images for precise style and content control.

Parameter Controls

Sliders for guidance scale (1.0–20.0), denoising steps (1–100), image dimensions, and seed for reproducibility.

Generation History Gallery

All generated images are saved with their prompts and parameters in a scrollable gallery.

Image Download

Download individual images or the full gallery as a ZIP file.

Prompt Enhancement

Optional LLM-powered prompt expansion to convert terse inputs into detailed, style-rich generation prompts.

Multiple Backend Support

Configure fal.ai, Replicate, Stability AI, or local Diffusers pipeline as the generation backend.

Side-by-Side Comparison

Run the same prompt with different parameters and compare outputs in a split view.

---

Tech Stack

Library / Tool	Role	Why This Choice
Diffusers / fal.ai SDK	Generation backend	Diffusion model inference
Streamlit / Gradio	Web UI	Interactive parameter controls and gallery
Pillow	Image handling	Image saving, resizing, format conversion
requests	API communication	Cloud generation API calls

Key packages detected in this repo: rembg · Pillow · streamlit

---

Getting Started

Prerequisites

• Python 3.9+ (or Node.js 18+ for TypeScript/JS projects)
• pip or npm package manager
• Relevant API keys (see Configuration section)

Installation

git clone https://github.com/Devanik21/BG-remover.git
cd BG-remover
pip install -r requirements.txt
python app.py

---

Usage

python app.py

# Or with Streamlit
streamlit run app.py

---

Configuration

Variable	Default	Description
API_KEY	(required)	API key for the configured generation backend
DEFAULT_MODEL	flux-schnell	Default generation model
DEFAULT_STEPS	20	Default inference steps

Copy .env.example to .env and populate all required values before running.

---

Project Structure

BG-remover/
├── README.md
├── requirements.txt
├── __init__.py
├── app.py
└── ...

---

Roadmap

• ControlNet support for structure-guided generation
• Style transfer and image-to-image mode
• Batch generation from CSV prompts
• Video generation integration
• User accounts with cloud gallery persistence

---

Contributing

Contributions, issues, and feature requests are welcome. Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/your-feature)
3. Commit your changes (git commit -m 'feat: add your feature')
4. Push to your branch (git push origin feature/your-feature)
5. Open a Pull Request

Please follow conventional commit messages and ensure any new code is documented.

---

Notes

API keys for the configured generation backend are required for cloud inference. Local inference requires a CUDA-capable GPU.

---

Author

Devanik Debnath
B.Tech, Electronics & Communication Engineering
National Institute of Technology Agartala

---

License

This project is open source and available under the MIT License.

---

Crafted with curiosity, precision, and a belief that good software is worth building well.