Cancelable Speaker Identification System Based on Optical-Like Encryption Algorithms

Biometric authentication is a rapidly growing trend that gaining increasing attention in the last decades. It achieves safe access to systems using biometrics instead of traditional passwords. The utilization biometric its original format makes it usable only once. Therefore, cancelable template should be used, so can replaced when attacked. Cancelable aims enhance security and privacy authentication. Digital encryption an efficient technique used order generate templates. In this paper, highly-secure algorithm proposed ensure secure data verification systems. considered paper speech signal. signal transformed into spectrogram. Then, spectrogram encrypted two cascaded optical algorithms. first Optical Scanning Holography (OSH) for efficiency as tool. OSH Double Random Phase Encoding (DRPE) by implementing Masks (RPMs). After algorithms, obtained. implemented through correlation estimation between enrolled test templates their format. If value larger than threshold value, user authorized. determined from genuine imposter distribution curves midpoint curves. implementation adopted software rather setup. illustrated simulation results. improve without deteriorating recognition accuracy. Simulation results give close-to-zero values Equal Error Rate (EER) close-to-one Area under Receiver Operator Characteristic (AROC) curve.