SIGNAL PROCESSING OF EEG WAVES USING CONVOLUTIONAL NEURAL NETWORK

Convolutional Neural Networks (CNNs) have been recently developed to be a stateof-the-art technology for critical applications such as ECG classification for patients, systemic health surveillance, power-electronic circuit malformation identification and motor defect detection. The raw signal can be used with CNNs directly. Moreover, it is feasible because of the easy and compact construction of such adaptive CNNs that only make linear transformations and low cost hardware. This document examines the main signal processing applications with a brief theoretical background of lightweight CNNs.