An ERP dataset for multi-information identity authentication

Conventional biometric authentication methods, such as fingerprint recognition, facial recognition, and voiceprint recognition, are vulnerable to spoofing, coercion, and data leakage. Electroencephalography (EEG), owing to its resistance to replication and involuntary nature, has shown considerable potential in secure biometric authentication, particularly in brain-computer interface (BCI) systems based on rapid serial visual presentation (RSVP). However, authentication relying solely on facial information remains limited by the single-dimensional nature of the authentication cue, insufficient resistance to attacks, and limited robustness in practical applications. Therefore, this study constructs, for the first time, a BIDS-compliant RSVP-EEG dataset for multi-information identity authentication, incorporating three types of authentication factors: target faces, target names, and verification images. Data were collected from 32 healthy participants, each of whom completed two experimental sessions for all three authentication tasks with an interval of >24 h. Identity-related and identity-unrelated stimuli were presented at a frequency of 10 Hz to elicit prominent recognition-related neural responses, and EEG signals were recorded using an eight-channel portable EEG system covering frontal and occipital regions. The proposed experimental design integrates multiple types and hierarchical levels of identity-related information, providing a solid data foundation for ERP analysis, RSVP-based cognitive experimental research, and the development of multi-information identity authentication systems.