gesture-recognition

Prediction of gestures based on capturing of muscles contractions and ML

Based on PyoMyo library.

plot-emg-data-mat.py

Capturing of RAW EMG data using myo armband and python matplotlib. Captured data are plotted in Matplotlib . 8 graphs are created - every graph for 1 armband sensor.

Capturing filtred EMG data using myo armband and python matplotlib. Captured data are plotted in Matplotlib . 8 graphs are created - every graph for 1 armband sensor. (Bandpass filter + rectified) (mode 0x01)

plot-emg-data-arr.py

RAW EMG data (array of 8 values)

Filtred EMG data(array of 8 values)(Bandpass filter + rectified) (mode 0x01)

emg-data-classifier.py

This script is dedicated to capture data from armband and learn (improve model) in real time.

It is using k-nearest neighbors algorithm.

Labbeled data are stored as array.

In order to delete all the data press E button on the keyboard.

Press numbers from 0 to 9 to label data.

https://youtu.be/xZ1mQtPdz5I

emg-data-classifier-anim2.py

This script is dedicated to capture data from armband, learn (improve model) in real time and show real-time hand animation.

(Start emg-data-classifier-anim.py, then in the project folder find folder HandAnim and run Hand.exe) (In hand animation was used Godot Engine) Script is using k-nearest neighbors algorithm.

Labbeled data are stored as array.

In order to delete all the data press E button on the keyboard.

Press numbers from 0 to 9 to label data.