Enhanced ECG-Based Biometric Authentication using a Hybrid CNN-LSTM Framework

Reliability for sensitive applications is limited by flaws in traditional biometric systems, such as fingerprint, face, and iris identification, which include ease of spoofing, replication, and degradation over time. Using a hybrid deep learning framework that combines Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNNs) provides a unique ECG-based biometric authentication system to address these problems. Enhanced security and accuracy are achieved by utilizing the unique and challenging-to-copy features of ECG signals. As the LSTM records temporal patterns and considerably enhances authentication efficiency, the CNN retrieves intricate features from unprocessed ECG data. As the experimental results show, the proposed CNNLNSTM framework outperforms conventional precision, recall, and error rates, achieving an impressive 98.5% accuracy. It’s a reliable option for real-time biometric authentication because of its high adaptability and scalability. Because it overcomes the drawbacks of earlier methods, the advanced technology provides a significant boost in biometric security.