ASE: Practical Acoustic Speed Estimation Beyond Doppler via Sound Diffusion Field

Sheng Lyu , Chenshu Wu

The University of Hong Kong

The official implementation of ASE: Practical Acoustic Speed Estimation Beyond Doppler via Sound Diffusion Field, accepted by ACM IMWUT / UbiComp 2025.

ASE vs. DFS. Walking speed $\vec{v}$ can be decomposed to radial speed $v_r$ and tangent speed $v_t$. DFS can only capture radial speed $v_r$, but fails to capture $v_t$. Conversely, ASE can capture both $v_r$ and $v_t$, a complete estimation of $v$.

---

Abstract

Passive human speed estimation plays a critical role in acoustic sensing. Despite extensive study, existing systems, however, suffer from various limitations: First, previous acoustic speed estimation exploits Doppler Frequency Shifts (DFS) created by moving targets and relies on microphone arrays, making them only capable of sensing the radial speed within a constrained distance. Second, the channel measurement rate proves inadequate to estimate high moving speeds. To overcome these issues, we present ASE, an accurate and robust Acoustic Speed Estimation system on a single commodity microphone. We model the sound propagation from a unique perspective of the acoustic diffusion field, and infer the speed from the acoustic spatial distribution, a completely different way of thinking about speed estimation beyond prior DFS-based approaches. We then propose a novel Orthogonal Time-Delayed Multiplexing (OTDM) scheme for acoustic channel estimation at a high rate that was previously infeasible, making it possible to estimate high speeds. We further develop novel techniques for motion detection and signal enhancement to deliver a robust and practical system. We implement and evaluate ASE through extensive real-world experiments. Our results show that ASE reliably tracks walking speed, independently of target location and direction, with a mean error of 0.13 m/s, a reduction of 2.5x from DFS, and a detection rate of 97.4% for large coverage, e.g., free walking in a 4m × 4m room. We believe ASE pushes acoustic speed estimation beyond the conventional DFS-based paradigm and will inspire exciting research in acoustic sensing.

---

Setup Instructions

To get started with ASE, follow these steps:

1. Clone the GitHub repository:

git clone https://github.com/aiot-lab/ASE.git
cd ASE

2. Make sure you have installed uv for virtual environment, and then run

uv sync

3. Activate the virtual environment

source .venv/bin/activate

Run the code

We provide two modes to run the code.

• UI Version: We use streamlit to build a webUI of ASE. You can use it to do the experiments and visualize the analysis results. To start the UI, run

  streamlit run app.py

  New recordings would be added to data folder and will be assigned with a unique index.
• Headless Analysis Version: To run the analysis in the headless mode, run

  python main.py analysis --config config_file/speed_gui.json --rec-idx 28

  Make sure you have the corresponding data in the data folder and input the rec-idx (a.k.a the index) you want to analyze.

Repository Structure

.
├── README.md
├── app.py
├── app_utils.py
├── audio
│   ├── Audio.py
│   ├── AudioProcessing.py
│   ├── Wave.py
│   ├── __init__.py
│   └── helpers.py
├── check_param.py
├── config_file
│   └── speed_gui.json
├── config_gui.py
├── data
│   ├── speed_Kasami_micArray RAW SPK_micArray RAW SPK_20250410-233950_28
│   └── speed_Kasami_micArray RAW SPK_micArray RAW SPK_20250410-234131_29
├── data_play
├── dataloader.py
├── fig
│   └── dfs_acf_all.png
├── index_manager.py
├── main.py
├── model
│   ├── ACF.py
│   ├── __init__.py
│   ├── acf_benchmark.py
│   ├── acf_wrapper.py
│   ├── alignment.py
│   ├── helpers.py
│   ├── optimized_acf.py
│   ├── save.py
│   └── timestamp.py
├── parser_config.py
├── pyproject.toml
├── utils.py
├── utils_gui.py
├── uv.lock
└── vis
    ├── __init__.py
    ├── vis_utils.py
    └── visual.py

---

Citation

@article{lyu2025ase,
  title={ASE: Practical Acoustic Speed Estimation Beyond Doppler via Sound Diffusion Field},
  author={Lyu, Sheng and Wu, Chenshu},
  journal={Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies},
  volume={9},
  number={3},
  pages={1--31},
  year={2025},
  publisher={ACM New York, NY, USA}
}

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.