Thursday, September 16, 2004
3:00-4:00 p.m.
226 Erickson Hall

ABSTRACT

The area of automatic speaker recognition has been dominated by systems using only short-term, low-level acoustic information, such as cepstral features. While these systems have produced reasonably low error rates, they ignore other levels of information beyond low-level acoustics that convey speaker information. This seminar will include a tutorial and focus on late-breaking research that exploits high-level information, e.g., idiosyncratic word usage and pronunciation, in automatic speaker recognition systems to add robustness and improve accuracy.

BIOGRAPHY

Joseph P. Campbell received B.S., M.S., and Ph.D. degrees in electrical engineering from Rensselaer Polytechnic Institute in 1979, The Johns Hopkins University in 1986, and Oklahoma State University in 1992, respectively. Joe is currently a Senior Member of the Technical Staff at MIT Lincoln Laboratory in the Information Systems Technology Group, where he conducts speech-processing research and specializes in advanced speaker recognition methods. Joe's current foci are high-level features for and forensic-style applications of speaker recognition, creating corpora to support speech processing research and evaluation, robust speech coding, biometrics, and cognitive radio. Before joining Lincoln, he served 22 years at the National Security Agency (NSA).

From 1979 to 1990, Dr. Campbell was a member of NSA's Narrowband Secure Voice Technology research group. Joe and his teammates developed the first DSP-chip software modem and LPC-10e, which enhanced the Federal Standard 1015 voice coder and improved US and NATO secure voice systems. He was the Principal Investigator and led the US Government's speech coding team in developing the CELP voice coder, which became Federal Standard 1016 and is the foundation of digital cellular and voice over the Internet telephony systems. From 1991 to 1998, Dr. Campbell was a senior scientist in NSA's Biometric Technology research group, where he led voice verification research. From 1994 to 1998, Joe chaired the Biometric Consortium, the US Government's focal point for research, development, test, evaluation, and application of biometric-based personal identification and verification technology. From 1998 to 2001, he led the Acoustics Section of NSA's Speech Research branch, conducting and coordinating research on and evaluation of speaker recognition, language identification, gender identification, and speech activity detection methods.

From 1991 to 1999, Dr. Campbell was an Associate Editor of the IEEE Transactions on Speech and Audio Processing. He was an IEEE Signal Processing Society Distinguished Lecturer in 2001. From 1991 to 2001, Joe taught Speech Processing at The Johns Hopkins University. Dr. Campbell is currently a member of the IEEE Signal Processing Society's Board of Governors; an Editor of Digital Signal Processing journal; a Chair of the International Speech Communication Association's Speaker and Language Characterization Special Interest Group (ISCA SpLC SIG); a member of ISCA, Sigma Xi, and the Acoustical Society of America; and a Senior Member of the IEEE.