BMT_EEG: A Novel EEG Dataset for evaluating the effects of new protocols on biometric authentication systems

Due to uniqueness, revocability, and resilience to theft, EEG biometric solutions are becoming more popular for securing high stake systems. Electroencephalography (EEG), a non-invasive method to record brain activity, has a crucial role in this domain. To propel research in EEG-based biometrics, we introduce a novel dataset: Biometric EEG Dataset (BMT_EEG), uniquely designed to address anti-spoof concerns of Biometric applications. The dataset captures neural responses of 20 subjects across three sessions, encompassing eleven protocols that include seven motor movement (MM) and motor imagery (MI) tasks, two visually evoked potentials (VEPs), and two baseline resting-state conditions. BMT_EEG provides a rich, multi-session view into task-related and baseline brain activity, enabling detailed analyses of neural signal patterns. Leveraging BMT_EEG, we developed a hybrid EEG_CNN-GRU model combining 1-D Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU) for modeling spatio-temporal dependencies for user authentication\/identification. The model was evaluated under both random and skilled forgery attacks in an inter-session setting, using two sessions for training and the third for testing, enhancing robustness against session variability and its suitability for real-world deployment scenario. The model achieved the best average Equal Error Rate (EER) of 1.46% for random forgery using the MI2 task and 7.01% for skilled forgery using the MI1 task. For identification, the model demonstrated an average precision of 74.29%, recall of 77.33%, and F1-score of 74.18%, demonstrating its effectiveness for EEG-based user recognition. This work underscores the importance of realistic evaluation protocols, comprehensive EEG datasets, and advanced deep learning architectures in advancing biometric research.