By Manas A. Pathak
This thesis discusses the privateness concerns in speech-based functions equivalent to biometric authentication, surveillance, and exterior speech processing companies. writer Manas A. Pathak offers suggestions for privacy-preserving speech processing purposes reminiscent of speaker verification, speaker id and speech attractiveness. the writer additionally introduces many of the instruments from cryptography and desktop studying and present thoughts for bettering the potency and scalability of the awarded strategies. Experiments with prototype implementations of the ideas for execution time and accuracy on standardized speech datasets also are incorporated within the textual content. utilizing the framework proposed may perhaps now guarantee that a surveillance service provider to hear for a recognized terrorist with out having the ability to pay attention dialog from non-targeted, blameless civilians.