EEG Person Identification with CNN GRU hybrid model

This project presents a state-of-the-art deep learning system for person identification using

Electroencephalography (EEG) signals. By leveraging a hybrid CNN-GRU architecture, the system achieved a 99.89% Top-1 accuracy across 109 subjects from the PhysioNet Motor

Movement/Imagery Database. The model is designed for real-time applications, boasting an

inference time of 15 milliseconds, making it a viable candidate for secure, high-speed biometric

authentication.

The system utilizes 10 specific motor cortex channels to

capture spatial-frequency features.