Toward EEG Based Biometric Authentication in VirtualReality: a Pilot Usability and Performance Study

Abstract

This pilot study investigates the feasibility and usability of EEG-based biometric authentication integrated into Virtual Reality environments. Leveraging consumer-grade EEG hardware (14-channel Emotiv EPOC-X), we collected short-duration EEG signals from 25 participants engaging in visually stimulated tasks within a controlled VR setting. Classification models including Ensemble Subspace K-Nearest Neighbors and Long Short-Term Memory networks achieved promising identification accuracies of up to 85% and 91.65%, respectively, demonstrating EEG's potential as a robust biometric modality for identity verification in immersive contexts. User experience evaluations indicate that the EEG-based system was well-received, showing high usability and minimal intrusiveness, with the majority of users expressing willingness to adopt such technology. Despite limitations related to the small sample size and homogeneous participant demographics, this study provides critical insights and identifies key considerations for future, more comprehensive research on EEG biometric authentication in Vitual Reality applications.