Skip to content

Latest commit

 

History

History
85 lines (69 loc) · 3.24 KB

README.md

File metadata and controls

85 lines (69 loc) · 3.24 KB

Gesture Recognition with mmWave Radar

Code repository for the paper: Towards Domain-independent and Real-time Gesture Recognition using mmWave Signal, IEEE TMC, 2022.

Authors: Yadong Li, Dongheng Zhang, Jinbo Chen, Jinwei Wan, Dong Zhang, Yang Hu, Qibin Sun, and Yan Chen

Affiliation: Intelligent Perception Lab, University of Science and Technology of China. Paper Link

System Overview

Introduction

  • This code repository mainly includes two parts described in the paper:

    • Signal Processing: Pre-processing the raw FMCW signal to get the dynamic range angle image (DRAI).

    • Deep Learning: A CNN-LSTM neural network that inputs DRAI and outputs gesture labels.

  • The code can be used to support other similar applications using wireless signals.

Signal processing
Neural Network

How to Access the Code

The USTC IP Lab has particular protocols for releasing the code and dataset. To access the code, please sign the code agreement, scan and send it to [email protected]. A notification email that includes the code will be sent within three days.

Installation and Demo

Signal processing

Setup:

conda create -n digesture python==3.8
conda activate digesture
pip install opencv-python matplotlib numpy pillow

Run the demo:

python signal_process_DRAI.py

Deep Learning

Setup:

conda activate digesture
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
pip install scikit-learn pandas tqdm

Train the network:

python train.py --data_path your_data_path_to_the_DRAI_dataset

Test the network:

python test.py --data_path your_data_path_to_the_DRAI_dataset

Citing

If you find this code useful for your research, please consider citing the following papers:

@ARTICLE{9894724,
  author={Li, Yadong and Zhang, Dongheng and Chen, Jinbo and Wan, Jinwei and Zhang, Dong and Hu, Yang and Sun, Qibin and Chen, Yan},
  journal={IEEE Transactions on Mobile Computing}, 
  title={Towards Domain-Independent and Real-Time Gesture Recognition Using mmWave Signal}, 
  year={2023},
  volume={22},
  number={12},
  pages={7355-7369},
  doi={10.1109/TMC.2022.3207570}}


@INPROCEEDINGS{10001175,
  author={Li, Yadong and Zhang, Dongheng and Chen, Jinbo and Wan, Jinwei and Zhang, Dong and Hu, Yang and Sun, Qibin and Chen, Yan},
  booktitle={GLOBECOM 2022 - 2022 IEEE Global Communications Conference}, 
  title={DI-Gesture: Domain-Independent and Real-Time Gesture Recognition with Millimeter-Wave Signals}, 
  year={2022},
  volume={},
  number={},
  pages={5007-5012},
  doi={10.1109/GLOBECOM48099.2022.10001175}}