Enhancing Biometric Security: A Machine Learning Approach to ECG-Based Authentication

Using ECG-based authentication, it offers a unique way to improve biometric security. By using a hybrid model that blends RNN and CNN, the system seeks to increase both resilience and accuracy. The procedure comprises preprocessing and gathering ECG data, then extensive feature extraction and model training. Impressively, 99.1% accuracy was attained by the suggested technique. With accuracy rates of 99.3% when at rest, 98.7% while modest exercise, and 95.6% in noisy situations, the system performed well when tested in a variety of scenarios. 98.2% accuracy, 97.8% precision, 98.5% recall, and 98.1% F1-score were the performance metrics of the CNN model; on the other hand, the RNN model yielded results of 97.5% accuracy, 96.9% precision, 97.8% recall, and 97.3% F1-score. The hybrid system scored 99.1% accuracy, 99.0% precision, 99.2% recall, and 99.1% F1-score CNN+RNN model, which beat the other two. This highlights the system's reliable operation and potential for useful biometric applications. Subsequent efforts will concentrate on augmenting the dataset, investigating supplementary deep learning architectures, and amalgamating multimodal biometric data.