The growing demand for secure and continuous authentication has highlighted the limitations of traditional biometric modalities such as passwords, fingerprints, and facial recognition. Electrocardiogram (ECG) signals have emerged as a robust biometric trait due to their physiological uniqueness, inherent liveness detection, and resistance to spoofing attacks. This work presents a secured multimodal biometric authentication system based on ECG signals and deep learning techniques. The proposed approach integrates signal preprocessing, heartbeat segmentation, and deep neural network-based feature learning within a multimodal fusion framework

Азаб М.А. (науч. рук. Коржук В.М.) A secured multimodal ecg biometric authentication system using deep learning // Сборник тезисов докладов конгресса молодых ученых. Электронное издание. – СПб: Университет ИТМО, [2026]. URL: https://kmu.itmo.ru/digests/article/16030