Wednesday, March 8, 2023
The Joint CARTE (University of Toronto) and University of Seoul Applied AI/DS seminar series welcomes Professor Ha-Jin Yu.
Registration: Please register through this link.
Abstract: Speaker recognition is the task of identifying who is speaking from their voices. Recently, deep learning has dramatically revolutionized speaker recognition. However, the performance degrades when the utterance is too short or noisy. We investigate the way of challenging such harsh conditions. Another threat to speaker verification is the deepfake using speech synthesis, conversion, or replaying recorded speech. It is also called spoofing. We devise anti-spoofing methods to ward off such attacks. This talk deals with current speech verification technology trends and our approaches.
Bio: HA-JIN YU received B.S., M.S., and Ph.D. degrees in computer science from KAIST, Korea, in 1990, 1992, and 1997, respectively. From 1997 to 2000, he was a Senior Researcher with LG Electronics. From 2000 to 2002, he was a Director with SL2 Ltd. Since 2002, he has been a Professor at the Computer Science Department, the University of Seoul. His research interests include speech and speaker recognition, acoustic scene classification, and machine learning. He is an Editor of The Journal of the Acoustical Society of Korea and the Journal of the Korean Society of Speech Sciences.