DARMS

DARMS is a Device-free Human Activity Recognition System based on Dual-channel Transformer Using WIFI Signals.

Requirements

MATLAB R2020a

PyTorch 1.10.0

Getting Started

1. Click here to download the raw CSI data and put it into the raw_data folder.

2. Using the python/mat2npy.py to genernate the dataset.

3. Then, you can quickly use DARMS by runing the python/DARMS_Main.py to train and validate the nerual network.

Project Structure

DARMS
├─ matlab                 // The pre-processsing algorithm programmed with matlab
├─ python                 // The nerual network programmed with python
├─ raw_data               // The raw CSI data
├─ tmp                    // Storing the intermediate results.
└─ dataset                // The dataset for training and validating the nerual network

Citation

If you feel our data, codes or paper helpful in your research, please cite this work.

@article{gu2022device, title={Device-Free Human Activity Recognition Based on Dual-Channel Transformer Using WiFi Signals}, author={Gu, Zhihao and He, Taiwei and Wang, Ziqi and Xu, Yuedong}, journal={Wireless Communications and Mobile Computing}, volume={2022}, publisher={Hindawi} }