S4P Summer School on Speech Signal Processing, Speaker Recognition and Diarization, DAIICT, 6-10 July 2019, Gandhinagar, India
Abstract: My first talk starts by outlining the requirements for successful biometric characteristics and explores whether these can be met with the use of voice recordings for speaker recognition. The remainder of the talk focuses upon one requirement in particular, namely that relating to the robustness of biometric systems to circumvention. Iwill discuss early work that explored security aspects of speaker recognition systems and more recent work which has begun to investigate the potential of the same technology to intrude on personal privacy. The final part of the talk touches upon recent regulation which places restrictions upon the extent to which voice recordings can be lawfully collected, stored and processed.