Evaluation of proposed modifications to MPE for large scale discriminative training

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

Volumn 4, Issue , 2007, Pages

  Povey, Daniel ^a   Kingsbury, Brian ^a  

^a IBM T J WATSON RESEARCH CENTER   (United States)

Author keywords

Discriminative training; Minimum bayes risk; Minimum divergence; Minimum phone error; Minimum phone frame error

Indexed keywords

CONVERGENCE OF NUMERICAL METHODS; DISCRIMINATORS; ERROR DETECTION; FUNCTION EVALUATION; TELEVISION BROADCASTING;

DISCRIMINATIVE TRAINING; MINIMUM BAYES RISK; MINIMUM DIVERGENCE; MINIMUM PHONE ERROR; MINIMUM PHONE FRAME ERROR (MPFE);

ACOUSTIC SIGNAL PROCESSING;

EID: 34547529083     PISSN: 15206149     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1109/ICASSP.2007.366914     Document Type: Conference Paper

References (9)

1. Povey D. and Woodland P.C., "Minimum Phone Error and I-smoothing for Improved Discriminative Training," in ICASSP, 2002.
2. Povey D., Discriminative Training for Large Voculabulary Speech Recognition, Ph.D. thesis, Cambridge University, 2004.
3. Soltau H., Kingsbury B., Mangu L., Povey D., Saon G., and Zweig G., "The IBM 2004 Conversational Telephony System for Rich Transcription in EARS," in ICASSP, 2005.
4. Du J., Liu P., Soong F.K., Zhou J-L., and Wang R-H., "Minimum Divergence Based Discriminative Training," in Interspeech, 2006.
5. Zheng J. and Stolcke A., "Improved Discriminative Training Using Phone Lattices," in Interspeech, 2005.
6. Gibson M. and Hain T., "Hypothesis Spaces For Minimum Bayes Risk Training In Large Vocabulary Speech Recognition," in Interspeech, 2006.
7. Soltau H., Saon G., Povey D., Mangu L., Kingsbury B., Omar M., and Zweig G., "The IBM 2006 GALE Arabic System," submitted to: ICASSP, 2007.
8. Povey D., Kingsbury B., Mangu L., Saon G., Soltau H., and Zweig G., "fMPE: Discriminatively trained features for speech recognition," in ICASSP, 2005.
9. H. Y. Chan and P. C. Woodland, "Improving broadcast news transcription by lightly supervised discriminative training," in Proc. ICASSP, 2004.