Home
Overall Project Goal
Our goal is to create a gesture recognition algorithm for purposes of interfacing sEMG sensors with personal devices and IoT. We believe that today the vast potential of EMG signal for providing convenient and natural way of device control is majorly overlooked. With our project we want to contribute to the field of human biosignals processing, which one day will set a new standards of human-computer communication!
Our Approach
In order to fulfill the aim of our project, we collected a dataset of sEMG signals derived from the forearm during the process of typing. We trained a classification neural network to recognize pushes on different buttons on the keyboard and implemented it within an Android application. We successfully showed that these subtle gestures could be recognized and translated into a set of characters.