Using MAP estimated parameters to improve HMM speech recognition performance

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

Volumn 1, Issue , 1994, Pages I229-I232

  Gotoh, Y ^a   Hochberg, M M ^b   Silverman, H F ^a  

^a BROWN UNIVERSITY   (United States)

^b UNIVERSITY OF CAMBRIDGE   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

HIDDEN MARKOV MODELS; MAXIMUM LIKELIHOOD ESTIMATION; PARAMETER ESTIMATION; SIGNAL PROCESSING;

DURATION MODELING; ESTIMATED PARAMETER; HIDDEN MARKOV MODELS (HMMS); MAP-BASED APPROACH; MAXIMUM A POSTERIORI; MAXIMUM LIKELIHOOD CRITERIA; PRIOR DISTRIBUTION; SPEECH RECOGNITION PERFORMANCE;

SPEECH RECOGNITION;

EID: 33744561188     PISSN: 15206149     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1109/ICASSP.1994.389313     Document Type: Conference Paper

References (9)

1. Stephen E. Levinson. Continuously variable duration hidden Markov models for automatic speech recognition. Computer Speech and Language, 1:29-45, 1986.
2. Michael M. Hochberg. A Comparison of State-Duration Modeling Techniques for Connected Speech Recognition. PhD thesis, Brown University, Providence, RI, 1993.
3. Michael M. Hochberg and Harvey F. Silverman. Constraining model duration variance in HMM-based connected-speech recognition. In Proceedings of Eurospeech-93, volume 1, pages 323-326, Berlin, September 1993.
4. Thomas H. Crystal and Arthur S. House. Segmental durations in connected-speech signals: Current results. Journal of the Acoustical Society of America, 83(4):1553-1573, April 1988.
5. James O. Berger. Statistical Decision Theory and Bayesian Analysis. Springer-Verlag, New York, 2nd edition, 1985.
6. Jean-Luc Gauvain and Chin-Hui Lee. MAP estimation of continuous density HMM: Theory and applications. In DARPA Workshop on Speech and Natural Language, pages 185-190, Harriman, NY, February 1992.
7. Jean-Luc Gauvain and Chin-Hui Lee. Improved acoustic modeling with Bayesian learning. In Proceedings of ICASSP-92, volume 1, pages 481-484, San Francisco, March 1992
8. Chin-Hui Lee and Jean-Luc Gauvain. Speaker adaptation based on MAP estimation of HMM parameters. In Proceedings of ICASSP-93, volume II, pages 558-561, Minneapolis, April 1993.
9. Harvey F. Silverman and Yoshihoko Gotoh. On the implementation and computation of training an HMM recognizer having explicit state durations and multiple-feature-set tied-mixture observation probabilities. Technical Report LEMS-129, Division of Engineering, Brown University, 1993.