Deep Learning AI Application to an EEG driven BCI Smart Wheelchair

Abstract

This paper describes designing, 3-D printing, building and testing of a prototype smart wheelchair using Emotive EPOC headset that will help the blind and paralyzed people who are unable to control parts of their body. It uses deep learning in order to recognize four different movements from the recorded EEG signal, to move left, right, forward, and stop. Data from 10 volunteers shows a success rate of 70% from the row EEG and 96% from the spectrum (Fourier Transform) of the data from the frequency bins corresponding to the delta, theta, alpha, beta, and gamma waves.