EEG-Based Person Authentication Method Using Deep Learning with Visual Stimulation

Abstract

Biometric authentication methods, such as fingerprint, used nowadays still have numerous potential security concerns and can be forged physically. To mitigate this issue, recent research has established non-physical signals such as brainwaves as viable sources for secured person authentication. As the brain signal has been found to be unique for individuals, electroencephalogram (EEG) signals can be potentially utilized as an identity discrimination tool, especially since EEG reflects the delicate difference in individuals' mental characteristics. However, the existing EEG-based authentication methods not only apply one or few techniques to stimulate the signals but also there are some vulnerabilities in the limited scope of the study. Therefore, we propose to combine steady-state visual evoked potential (SSVEP) and event-related potential (ERP) features to discriminate the distinction between individuals and apply the Long Short-Term Memory (LSTM) network for the analysis. The proposed methodology is divided into three stages. Firstly, we collect raw EEG data from the 20 human subjects, whose brains were stimulated by 7.5 Hz square SSVEP with targeted and non-targeted Snodgrass-Vanderwart's set of images as ERP stimulus. Then, the raw data is pre-processed by a notch filter, band-pass filter, and eye blink artifacts removal. The LSTM architecture is used to process data of individuals for predicting the results. After prediction, the performance is evaluated based on False Acceptance Rates and False Rejection Rates. The EEG-based authentication method with visual stimulus demonstrates high verification accuracy and can be applied with some improvement in the future through the use of brain connectivity techniques.