Improving robustness of MLLR adaptation with speaker-clustered regression class trees

Computer Speech and Language

Volumn 23, Issue 2, 2009, Pages 176-199

  Mandal, Arindam ^a   Ostendorf, Mari ^a   Stolcke, Andreas ^b,c  

^a University of Washington   (United States)

^b SRI INTERNATIONAL   (United States)

^c INTERNATIONAL COMPUTER SCIENCE INSTITUTE   (United States)

Author keywords

Regression class trees; Speaker adaptation; Speaker clustering; Speech recognition

Indexed keywords

BOOLEAN FUNCTIONS; ERROR ANALYSIS; LEARNING SYSTEMS; MATHEMATICAL TRANSFORMATIONS; MAXIMUM LIKELIHOOD ESTIMATION; REGRESSION ANALYSIS; TREES (MATHEMATICS);

AUTOMATIC SPEECH RECOGNITION SYSTEMS; CLUSTERED REGRESSION; CLUSTERING PROCEDURES; EIGEN SPACE; LINEAR COMBINATION; MAXIMUM-LIKELIHOOD LINEAR REGRESSION; MLLR ADAPTATION; PERFORMANCE LOSSES; RECOGNITION PERFORMANCE; REGRESSION CLASS TREES; SIGNIFICANT REDUCTION; SPEAKER ADAPTATION; SPEAKER CLUSTERING; SPEAKER VARIABILITY; TREE STRUCTURES; UNSUPERVISED ADAPTATION; WORD ERROR RATES;

SPEECH RECOGNITION;

EID: 53849127143     PISSN: 08852308     EISSN: 10958363     Source Type: Journal    
DOI: 10.1016/j.csl.2008.05.004     Document Type: Article

References (38)

1. Anastasakos, T., McDonough, J., Makhoul, J., 1997. Speaker adaptive training: a maximum likelihood approach to speaker normalization. In: Proc. of ICASSP, vol. 2, pp. 1043-1046.
2. Bocchieri, E., Digalakis, V., Corduneanu, A., Boulis, C., 1999. Correlation modeling of MLLR transform biases for rapid HMM adaptation to new speakers. In: Proc. of ICASSP, vol. 2, pp. 773-776.
3. Boulis C., Diakoloukas V., and Digalakis V. Maximum likelihood stochastic transformations adaptation for medium and small data sets. Computer Speech & Language 15 3 (2001) 257-287
4. Chen, K.T., Liau, W.W., Wang, H.M., Lee, L.S., 2000. Fast speaker adaptation using eigenspace-based maximum likelihood linear regression. In: Proc. of ICSLP, vol. III, pp. 742-745.
5. Cieri, C., Miller, D., Walker, K., 2004. The Fisher corpus: a resource for the next generations of speech-to-text. In: Fourth International Conference on Language Resources and Evaluation.
6. Dempster A., Laird N., and Rubin D. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society 39 1 (1977) 1-38
7. Digalakis V., Rtischev D., and Neumeyer L. Speaker adaptation using constrained estimation of Gaussian mixtures. IEEE Transactions on Speech and Audio Processing 3 5 (1995) 357-366
8. Ferrer, L., Sönmez, K., Kajarekar, S., 2005. Class-based score combination for speaker recognition. In: Proc. of Eurospeech, pp. 2173-2176.
9. Gales, M., 1996. The generation and use of regression class trees for MLLR adaptation. Tech. Rep. CUED/F-INFENG/TR263, Cambridge University.
10. Gales, M., 1997. Transformation smoothing for speaker and environmental adaptation. In: Proc. of Eurospeech, vol. 4, pp. 2067-2070.
11. Gales M. Maximum likelihood linear transformations for HMM-based speech recognition. Computer Speech & Language 12 (1998) 75-98
12. Gales M. Cluster adaptive training of hidden Markov models. IEEE Transactions on Speech and Audio Processing 8 4 (2000) 417-428
13. Gales M., and Woodland P. Mean and variance compensation within the MLLR framework. Computer Speech & Language 10 (1996) 249-264
14. Haeb-Umbach R. Automatic generation of phonetic regression class trees for MLLR adaptation. IEEE Transactions on Speech and Audio Processing 9 3 (2001) 299-302
15. Huang, C., Chen, T., Li, S., Chang, E., Zhou, J., 2001. Analysis of speaker variability. In: Proc. of Eurospeech, vol. 2, pp. 1377-1380.
16. Hwang, M.-Y., Huang, X., 1998. Dynamically configurable acoustic models for speech recognition. In: Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP'98, vol. 2, pp. 669-672.
17. Hwang, M.-Y., Lei, X., Wang, W., Shinozaki, T., 2006. Investigation on Mandarin broadcast news speech recognition. In: Proc. of ICSLP, pp. 1233-1236.
18. Imamura, A., 1991. Speaker adaptive HMM-based speech recognition with a stochastic speaker classifier. In: Proc. of ICASSP, vol. 2, pp. 841-844.
19. Kannan A., Ostendorf M., and Rohlicek J.R. Maximum likelihood clustering of gaussians for speech recognition. IEEE Transactions in Speech and Audio Processing 2 3 (1994) 453-455
20. Kosaka, T., Sagayama, S., 1994. Tree structured speaker clustering for fast speaker adaptation. In: Proc. of ICASSP, vol. 1, pp. 245-248.
21. Kuhn R., Junqua J., Nguyen P., and Niedzielski N. Rapid speaker adaptation in eigenvoice space. IEEE Transactions on Speech and Audio Processing 8 6 (2000) 695-707
22. Labov, W., 1996. The organization of dialect diversity in North America. In: Fourth International Conference on Spoken Language Processing.
23. Leggetter, C., 1995. Improved acoustic modelling for HMMs using linear transformations. Ph.D. Thesis, University of Cambridge.
24. Leggetter C., and Woodland P. Maximum likelihood linear regression for speaker adaptation of HMMs. Computer Speech & Language 9 (1995) 171-185
25. Mak, B., Hsiao, R., 2004. Improving eigenspace-based MLLR adaptation by kernel PCA. In: Proc. of ICSLP, vol. I, pp. 13-16.
26. Mandal, A., Ostendorf, M., Stolcke, A., 2005. Leveraging speaker-dependent variation of adaptation. In: Proc. of Eurospeech, pp. 1793-1796.
27. Mandal, A., Ostendorf, M., Stolcke, A., 2006. Speaker clustered regression-class trees for MLLR adaptation. In: Proc. of ICSLP, pp. 1133-1136.
28. Mardia K., Kent J., and Bibby J. Multivariate Analysis (1979), Academic Press
29. Padmanabhan M., Bahl L., Nahamoo D., and Picheny M. Speaker clustering and transformation for speaker adaptation in speech recognition systems. IEEE Transactions on Speech and Audio Processing 6 1 (1998) 71-77
30. R Development Core Team, 2005. R: a language and environment for statistical computing. R Foundation for Statistical Computing, ISBN 3-900051-07-0. .
31. Sankar, A., Beaufays, F., Digilakis, V., 1995. Training data clustering for improved speech recognition. In: Proc. of Eurospeech, vol. 1, pp. 502-505.
32. Sankar, A., Neumeyer, L., Weintraub, M., 1996. An experimental study of acoustic adaptation algorithms. In: Proc. of ICASSP, vol. 2, pp. 713-716.
33. Sankar, A., Gadde, R., Weng, F., 1999. SRI's 1998 broadcast news system - towards faster, smaller, and better speech recognition. In: DARPA Broadcast News Workshop, pp. 281-286.
34. Stolcke, A., Ferrer, L., Kajarekar, S., Shriberg, E., Venkataraman, A., 2005. MLLR transforms as features in speaker recognition. In: Proc. of Eurospeech, pp. 2425-2428.
35. Stolcke A., Chen B., Franco H., Gadde R., Graciarena M., Hwang M.Y., Kirchoff K., Lei X., Mandal A., Morgan N., Ng T., Ostendorf M., Sonmez K., Venkataraman A., Vergyri D., Wang W., Zheng J., and Zhu Q. Recent innovations in speech-to-text transcription at SRI-ICSI-UW. IEEE Transactions on Audio, Speech and Language Processing 14 5 (2006) 1729-1744
36. Venkataraman, A., Stolcke, A., Wang, W., Vergyri, D., Gadde, V., Zheng, J., 2004. SRI's 2004 broadcast news speech to text system. In: EARS RT04 Workshop.
37. Woodland, P., Gales, M., Pye, D., 1996. Improving environmental robustness in large vocabulary speech recognition. In: Proc. of ICASSP, vol. 1, pp. 65-68.
38. Young, S., Odell, J., Woodland, P., 1994. Tree based state tying for high accuracy modelling. In: Proc. ARPA Spoken Language Technology Workshop, pp. 405-410.