Key points are not available for this paper at this time.
The electroencephalogram (EEG) biometric-based personal identification system gained popularity due to its advantages. It is also difficult to replicate compared to traditional biometrics such as fingerprints, facial features, or voice, which make it a suitable candidate for the future personal identification system. However, the recent study of EEG as a biometric modality of identification system is still leaving issues. They still rely on high-resolution data, long acquisition times, and manually extracted features. This condition may not be suitable in practice. Therefore, in this study, we propose a novel single-channel EEG personal identification system. The proposed system takes 1 second of EEG signal from the P8 channel. The input signal is then evaluated using a customized convolutional neural network (CNN) model. The proposed system obtains a promising result with an accuracy of 98.95%. It has comparable performance to previous studies, although through a single channel and a short-length EEG signal.
Hendrawan et al. (Thu,) studied this question.