April 14, 2005 01 h 01 min
April 14, 2005 24 min
May 12, 2005 52 min
February 4, 2005 01 h 18 min
October 17, 2007 49 min
June 27, 2007 01 h 12 min
July 11, 2007 48 min
September 12, 2007 01 h 07 min
September 19, 2007 01 h 13 min
September 26, 2007 01 h 00 min
October 3, 2007 01 h 12 min
October 10, 2007 01 h 10 min
October 24, 2007 50 min
November 21, 2007 57 min
0:00/0:00
Malcolm SLANEY, invité par l'équipe Représentation Musicale,
présente en anglais :
"Pitch-Gesture Modeling Using Subband Autocorrelation Change Detection"
Calculating speaker pitch (or f0) is typically the first computational step in modeling tone and intonation for spoken language understanding. Usually pitch is treated as a fixed, single-valued quantity. The inherent ambiguity judging the octave of pitch, as well as spurious values, leads to errors in modeling pitch gestures that propagate in a computational pipeline.
We present an alternative that instead measures changes in the harmonic structure using a subband autocorrelation change detector (SACD).
This approach builds upon new machine-learning ideas for how to integrate autocorrelation information across subbands. Importantly however, for modeling gestures, we preserve multiple hypotheses and integrate information from all harmonics over time. The benefits of SACD over standard pitch approaches include robustness to noise and amount of voicing. This is important for real-world data in terms of both acoustic conditions and speaking style.
We discuss applications in tone and intonation modeling, and demonstrate the efficacy of the approach in a Mandarin Chinese tone-classification experiment. Results suggest that SACD could replace conventional pitch-based methods for modeling gestures in selected spoken-language processing tasks.
Biography
Malcolm Slaney (Fellow, IEEE) is a Principal Scientist in Microsoft Research's Conversational Systems Research Center in Mountain View, CA.
Before that he held the same title at Yahoo! Research, where he worked on multimedia analysis and music- and image-retrieval algorithms in databases with billions of items. He is also a (consulting) Professor at Stanford University's Center for Computer Research in Music and Acoustics (CCRMA), Stanford, CA, where he has led the Hearing Seminar for the last 20 years.
Before Yahoo!, he has worked at Bell Laboratory, Schlumberger Palo Alto Research, Apple Computer, Interval Research, and IBM's Almaden Research Center. For the last several years he has helped lead the auditory and attention groups at the NSF-sponsored Telluride Neuromorphic Cognition Workshop. He is a coauthor, with A. C. Kak, of the IEEE book Principles of Computerized Tomographic Imaging. This book was republished by SIAM in their Classics in Applied Mathematics series. He is coeditor, with S. Greenberg, of the book Computational Models of Auditory Function.
Prof. Slaney has served as an Associate Editor of the IEEE TRANSACTIONS ON AUDIO, SPEECH, AND SIGNAL PROCESSING, IEEE MULTIMEDIA MAGAZINE, the PROCEEDINGS OF THE IEEE, and the ACM Transactions on Multimedia Computing, Communications, and Applications.
Malcolm SLANEY, invité par l'équipe Représentation Musicale,
présente en anglais :
"Pitch-Gesture Modeling Using Subband Autocorrelation Change Detection"
Calculating speaker pitch (or f0) is typically the first computational step in modeling tone and intonation for spoken language understanding. Usually pitch is treated as a fixed, single-valued quantity. The inherent ambiguity judging the octave of pitch, as well as spurious values, leads to errors in modeling pitch gestures that propagate in a computational pipeline.
We present an alternative that instead measures changes in the harmonic structure using a subband autocorrelation change detector (SACD).
This approach builds upon new machine-learning ideas for how to integrate autocorrelation information across subbands. Importantly however, for modeling gestures, we preserve multiple hypotheses and integrate information from all harmonics over time. The benefits of SACD over standard pitch approaches include robustness to noise and amount of voicing. This is important for real-world data in terms of both acoustic conditions and speaking style.
We discuss applications in tone and intonation modeling, and demonstrate the efficacy of the approach in a Mandarin Chinese tone-classification experiment. Results suggest that SACD could replace conventional pitch-based methods for modeling gestures in selected spoken-language processing tasks.
Malcolm Slaney (Fellow, IEEE) is a Principal Scientist in Microsoft Research's Conversational Systems Research Center in Mountain View, CA.
Before that he held the same title at Yahoo! Research, where he worked on multimedia analysis and music- and image-retrieval algorithms in databases with billions of items. He is also a (consulting) Professor at Stanford University's Center for Computer Research in Music and Acoustics (CCRMA), Stanford, CA, where he has led the Hearing Seminar for the last 20 years.
Before Yahoo!, he has worked at Bell Laboratory, Schlumberger Palo Alto Research, Apple Computer, Interval Research, and IBM's Almaden Research Center. For the last several years he has helped lead the auditory and attention groups at the NSF-sponsored Telluride Neuromorphic Cognition Workshop. He is a coauthor, with A. C. Kak, of the IEEE book Principles of Computerized Tomographic Imaging. This book was republished by SIAM in their Classics in Applied Mathematics series. He is coeditor, with S. Greenberg, of the book Computational Models of Auditory Function.
Prof. Slaney has served as an Associate Editor of the IEEE TRANSACTIONS ON AUDIO, SPEECH, AND SIGNAL PROCESSING, IEEE MULTIMEDIA MAGAZINE, the PROCEEDINGS OF THE IEEE, and the ACM Transactions on Multimedia Computing, Communications, and Applications.