System and method for automatic synthetic speech detection for speech based biometric authentication

Abstract

The present invention discloses a system and method for automatic systematic speech detection for speech based biometric authentication. The present automatic systematic speech detection comprises recording speech signal sample from user end, pre-processing the speech signal sample to emphasize the said speech signal sample and converting into finite number of short segment speech signal samples, extracting feature of the pre-processed speech signal sample by involving inverted speech-signal-based overlapped block transformation (ISOBT), providing created target models from natural and synthetic speech data respectively and detecting synthetic speech signal by computing likelihood score between the feature of the speech signal sample and said created target model of said natural and synthetic speech data.