PythonMyoLinux

This repository contains the following:

1. A Python (3.5) package to collect sEMG and IMU measurements from Myo armband devices, featuring:
   1. An implementation of the Bluegiga API (BGAPI)
   2. Use of GAP and GATT client to:
      • Discover, and connect to devices
      • Search for and subscribe to available services
   3. Use of Myo specific commands

 

2. A GUI demonstration

   1. Overall system

   2. GUI tabs

      1. Data Collection
      2. Online Training
      3. Online Prediction

 

The GUI demonstration currently supports up to two Myo armband devices over a BLED112 Bluetooth dongle(s), and is intended to be used on a Linux distribution.

The GUI allows for collection of data with (crude) ground truth, as well as online training/testing. For refining of ground truth, and offline training/testing, please see the NinaTools respository.

 

Setup & Usage

First install the pymyolinux package

git clone https://github.com/sebastiankmiec/PythonMyoLinux.git
cd PythonMyoLinux
pip install .

 

1. Pymyolinux package demonstration

conda install -c conda-forge pyserial -y
python pymyolinux_example.py

 

2. GUI demonstration

Install remaining dependencies (first setup the NinaTools respository)

conda install -c conda-forge/label/cf201901 pyqt=5.6.0 pyqtgraph=0.10.0 -y

Distribution specific steps (for CentOS 7 only), to allow video playback

yum install gstreamer1.x86_64 gstreamer1-libav.x86_64 gstreamer1-plugins-bad-freeworld.x86_64 gstreamer1-plugins-good.x86_64 
yes | cp /lib64/gstreamer-1.0/* /path_to_miniconda/envs/testnina/lib/gstreamer-1.0/

To finally run the GUI demonstration

python gui_demo/gui_main.py

 

References

1. "Getting Started with Bluetooth Low Energy" by O'Reilly Media, Inc.
   • For a brief understanding of the Bluetooth stack, and Bluetooth core specification.
2. https://www.silabs.com/products/wireless/bluetooth/bluetooth-low-energy-modules/bled112-bluetooth-smart-dongle
   • See "Bluetooth Smart Software API Reference Manual for BLE Version 1.7", to understand BGAPI.
3. https://github.com/mjbrown/bgapi
   • A major inspiration for this work.
4. https://github.com/thalmiclabs/myo-bluetooth
   • Contains Myo specific commands.