The Impact of Data Segmentation Parameters on Performance in ECG-Based Identify Recognition Systems

This study focuses on the segment duration required for identity recognition systems using electrocardiogram (ECG) signals, which are widely employed in disease diagnostics. Signals from the ECG-ID dataset were preprocessed and evaluated without detecting any critical points. The signals were segmented based on different parameters and fed into a Convolutional Neural Network (CNN) model, with results analyzed accordingly. The findings indicate that successful identity recognition can be achieved even with short segment durations. This highlights significant potential for developing more efficient and faster solutions in biometric security and identity verification systems.