Authentication System Based on Event-related Potentials Using AR Glasses

Abstract

Although conventional biometric authentication methods have been developed for identification, they still involve some problems with hacking, copying, and theft. In this study, we propose a biometric authentication system based on electroencephalography (EEG) using the rapid serial visual presentation (RSVP) paradigm with stimuli of photographs of people displayed on augmented reality (AR) glasses. The event-related potentials (ERPs) of subjects were detected while they were exposed to two types of photographs, one that they were familiar with and the other that they were not. The amplitude of P3a, P3b and late positive potential (LPP) was smaller when subjects were exposed to familiar photographs compared to photographs of unfamiliar people. However, there was no change in the latency of P3a, P3b and LPP. Finally, the performance of the system was evaluated in terms of false rejection rate (FRR) and false acceptance rate (FAR). Thus, we identified the unique potential of EEG as a biometric authentication method by analyzing the response of ERPs during the RSVP paradigm.