ANN and Deep Learning Classifiers for BCI applications
K. Aswin, Manav Purushothaman, Polisetty Sritharani, T. S. Shiny Angel
Abstract
The Brain computer interface (BCI) or neural control interface is a technology that allows humans to control a computer or other computing devices on the basis of information inferred from thoughts. The communication is between a wired brain and an external device. The study comprises the acquisition of EEG signals and its classification. The classification process of EEG signals includes signal detection, feature extraction and classification of brain waves. The classification works on the principle of finding the best hyper-plane that separates the two classes in input space. In the proposed work, the application of brain waves for the direction control of a wheelchair in forward and backward direction is studied. Developed artificial neural network and deep neural network for the classification of EEG signals, and compared their performances using precision, recall and f1-score. Also a website application was developed for the advanced prediction and control.