This repo contains the code for the experiments in the papers:

1. Yu Hu, Yongkang Wong, Wentao Wei, Yu Du, Mohan Kankanhalli, Weidong Geng. " A novel attention-based hybrid CNN-RNN architecture for sEMG-based gesture recognition"
2. Yu Hu, Yongkang Wong, Qingfeng Dai, Mohan Kankanhalli, Weidong Geng. " sEMG-based gesture recognition with embedded virtual hand poses and adversarial learning"

• A CUDA compatible GPU
• Ubuntu >= 14.04 or any other Linux/Unix that can run Docker
• Docker
• Nvidia Docker

• Pull docker image for the first paper

  docker pull zjucapg/semg:latest
  

• Pull docker image for the second paper

  docker pull zjucapg/semgtf:latest  or  docker pull registry.cn-hangzhou.aliyuncs.com/semgtf/semgtf:latest
  

• Dataset

  Eleven databases including Ninapro DB1-DB7, CapgMyo DBa-DBc and CSL-HDEMG can be used for training and test.

  mkdir .cache
  # put NinaPro DB1 in .cache/ninapro-db1 or NinaPro DB7 in .cache/ninapro-db7
  # put CapgMyo DB-a in .cache/dba or DB-b in .cache/dbb or DB-c in .cache/dbc
  # put CSL-HDEMG in .cache/csl
  

  The NinaPro DB1 needs to be segmented by gesture labels and stored in Matlab format as follows..cache/ninapro-db1/data/sss/ggg/sss_ggg_ttt.mat contains a field data reprensents the trial ttt of gesture ggg of subject sss. And numbers start from zero. Gesture 0 is the rest gesture.

  For instance, .cache/ninapro-db1/data/000/001/000_001_000.mat is the 0th trial of 1st gesture of the 0th subject.

  You can download the original dataset from https://www.idiap.ch/project/ninapro/database or download the prepared dataset from our site http://zju-capg.org/myo/data/ninapro-db1.zip. CapgMyo and CSL-HDEMG datasets can be acquired on http://zju-capg.org/myo/data and http://www.csl.uni-bremen.de/cms/forschung/bewegungserkennung, respectively.

• Quick Start

  # Get into the capg/semg:mscnn container
  nvidia-docker run -ti -v your_projectdir:/code zjucapg/semg /bin/bash
  # first paper
  # Train
  sh scripts/exp.sh
  # Test
  python scripts/test.py
  
  # second paper
  # Train
  sh exp.sh
  # Test
  sh test.sh
  

Licensed under an GPL v3.0 license.

@article{hu2018novel,
  title={A novel attention-based hybrid CNN-RNN architecture for sEMG-based gesture recognition},
  author={Hu, Yu and Wong, Yongkang and Wei, Wentao and Du, Yu and Kankanhalli, Mohan and Geng, Weidong},
  journal={PloS one},
  volume={13},
  number={10},
  pages={e0206049},
  year={2018},
  publisher={Public Library of Science}
}
@article{hu2019semg,
  title={sEMG-Based Gesture Recognition With Embedded Virtual Hand Poses and Adversarial Learning},
  author={Hu, Yu and Wong, Yongkang and Dai, Qingfeng and Kankanhalli, Mohan and Geng, Weidong and Li, Xiangdong},
  journal={IEEE Access},
  volume={7},
  pages={104108--104120},
  year={2019},
  publisher={IEEE}
}