A Convolutional Neural Network based on Batch Normalization and Residual Block for P300 Signal Detection of P300-speller System

摘要

How to detect P300 signal efficiently and accurately is great importance to improve the performance of P300-spe11er system, one of a type brain-computer interface (BCI) system. In present study, we proposed a novel convolutional neural network (CNN) for P300 signal detection of P300-spe11er system based on traditional CNN, batch normalization, and residual block, which can extract the feature of P300 signal from spatial and temporal domain with little preprocessing. We compared the results of P300 signal detection and character recognition between traditional CNN and novel CNN, and the results showed that the proposed CNN got greater accuracy of P300 detection and character recognition than traditional CNN.