An Interactive System for the Study of EMG Signal ML-Processing and Prototyping of Human-Machine Interfaces

Abstract

A system has been created for controlling devices using EMG signals from a small number of sensors by recognizing signal sections corresponding to the operator's gestures. The solution implements a scheme of three independent components: sensors, a user interface and a data processor, connected by a simple data transfer protocol, which allows replacing any of the parts if necessary. Software has the realization of the most popular ML-algorithms, which widely used for EMG gesture recognition and intuitive GUI. It allows to use machine learning algorithms to users without any special background.