2008年清华大学信息技术研究院系列学术报告6

Biography

Michael Picheny is the Senior Manager of the Speech and Language Algorithms Group at the IBM TJ Watson Research Center . He has worked in the Speech Recognition area since 1981, joining IBM after finishing his doctorate at MIT. He has been heavily involved in the development of almost all of IBM's recognition systems, ranging from the world's first real-time large vocabulary discrete system through IBM's current product lines for telephony and embedded systems. He is the co-holder of over 20 patents, served as the chairman of the Speech Technical Committee of the IEEE Signal Processing Society from 2002-2004, and is a Fellow of the IEEE. He is currently a member of the board of ISCA (International Speech Communication Association).

Abstract

Large Vocabulary Speech Recognition systems have been deployed for over ten years by IBM and other speech technology providers in a variety of limited applications. However, achieving high accuracy across a variety of task domains, accents, channels, and speaking styles still remains a research challenge to the technical community. This talk will describe the state-in-the-art of Large Vocabulary Speech Recognition and also present some of the more recent techniques developed at IBM Research to achieve accuracy improvements. Advances in discriminative training, language modeling, and novel single-channel speaker separation techniques will be specifically highlighted, as well as recent work in the closely related application area of Spoken Term Detection.