Multimodal Biometric Authentication for VR/AR using EEG and Eye Tracking

Electroencephalogram (EEG) signals can enable an additional non-intrusive input modality especially when paired with a wearable headset (i.e. AR/VR). A great challenge in using EEG data for Brain-Computer Interface (BCI) algorithms is its poor generalization performance across users. Taking advantage of these inter-user differences, we investigate the potential in using this technology for user authentication – similar to facial recognition in smartphones. Additionally, we evaluate this in combination with eye tracking data which is also readily available in such headsets. We develop a biometric authentication systems for each of these systems and for their fusion. We formulate a novel evaluation paradigm using publicly available EEG motor imagery and eye tracking data and demonstrate strong feasibility towards using EEG and eye tracking for authentication.