Training data selection for improving discriminative training of acoustic models

Pattern Recognition Letters

Volumn 30, Issue 13, 2009, Pages 1228-1235

  Chen, Berlin ^a   Liu, Shih Hung ^a   Chu, Fang Hui ^a  

^a NATIONAL TAIWAN NORMAL UNIVERSITY   (Taiwan)

Author keywords

Acoustic models; Continuous speech recognition; Data selection; Discriminative training; Entropy; Phone accuracy

Indexed keywords

ACOUSTIC MODEL; ACOUSTIC MODELS; BROADCAST NEWS; DATA SELECTION; DISCRIMINATIVE ACOUSTIC MODEL; DISCRIMINATIVE TRAINING; GAUSSIANS; LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION; PHONE ACCURACY; POSTERIOR PROBABILITY; SPEECH TRANSCRIPTIONS; TRAINING DATA;

ACOUSTICS; CONTINUOUS SPEECH RECOGNITION; ENTROPY; SPEECH TRANSMISSION; TELEPHONE SETS; TRANSCRIPTION; TURNAROUND TIME;

DATA REDUCTION;

EID: 68149178821     PISSN: 01678655     EISSN: None     Source Type: Journal    
DOI: 10.1016/j.patrec.2009.05.009     Document Type: Article

References (35)

1. Aubert X.L. An overview of decoding techniques for large vocabulary continuous speech recognition. Comput. Speech Language 16 (2002) 89-114
2. Bahl, L.R., Brown, P.F., de Souza, P.V., Mercer, R.L., 1986. Maximum mutual information estimation of hidden Markov model parameters for speech recognition. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp. 49-52.
3. Chan, H.Y., Woodland, P.C., 2004. Improving broadcast news transcription by lightly supervised discriminative training. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp.737-740.
4. Chen, B., Kuo, J.W., Tsai, W.H., 2004. Lightly supervised and data-driven approaches to mandarin broadcast news transcription. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp. 777-780.
5. Chiu, H.S., Chen, B., 2007. Word topical mixture models for dynamic language model adaptation, In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp. 169-172.
6. Gales M.J.F. Maximum likelihood multiple subspace projections for hidden Markov models. IEEE Trans. Speech Audio Process. 10 2 (2002) 37-47
7. Jiang H., Soong F.K., and Lee C.H. A dynamic in-search data selection method with its applications to acoustic modeling and utterance verification. IEEE Trans. Speech Audio Process. 13 5 (2005) 945-955
8. Jiang H., Li X.W., and Liu C.J. Large margin hidden Markov models for speech recognition. IEEE Trans. Audio, Speech Language Process. 14 5 (2006) 1584-1595
9. Juang B.H., Chou W., and Lee C.H. Minimum classification error rate methods for speech recognition. IEEE Trans. Speech Audio Process. 5 3 (1997) 257-265
10. Kumar, N., 1997. Investigation of Silicon-Auditory Models and Generalization of Linear Discriminant Analysis for Improved Speech Recognition. Ph.D. Dissertation, John Hopkins University.
11. Kuo J.W., Liu S.H., Wang H.M., and Chen B. An empirical study of word error minimization approaches for mandarin large vocabulary speech recognition. Internat. J. Comput. Linguistic Chinese Language Process. 11 3 (2006) 201-222
12. Lee, H.S., Chen, B., 2009. Empirical error rate minimization based linear discriminant analysis. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp. 1801-1804.
13. Li, J., 2008. Soft Margin Estimation for Automatic Speech Recognition. Ph.D. Dissertation, Georgia Institute of Technology.
14. Li J., Ma B., and Lee C.H. Approximate test risk bound minimization through soft margin estimation. IEEE Trans. Audio, Speech Language Process. 15 8 (2007) 2393-2404
15. Liu, S.H., Chu, F.H., Lin, S.H., Chen, B., 2007a. Investigating data selection for minimum phone error training of acoustic models. In: Proc. IEEE Internat. Conf. on Multimedia and Expo, pp. 348-351.
16. Liu, S.H., Chu, F.H., Lin, S.H., Lee, H.S., Chen, B., 2007b. Training data selection for improving discriminative training of acoustic models. In: Proc. IEEE Workshop on Automatic Speech Recognition and Understanding, pp. 284-289.
17. Liu S.H., Chu F.H., Lo Y.T., and Chen B. Improved minimum phone error based discriminative training of acoustic models for Mandarin large vocabulary continuous speech recognition. Internat. J. Comput. Linguistics Chinese Language Process. 3 (2008) 327-342
18. Mathias, L., Yegnanarayanan, G., Fritsch, J., 2005. Discriminative training of acoustic models applied to domains with unreliable transcripts. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp. 109-112.
19. McDermott E., Hazen T.J., Roux J.L., Nakamura A., and Katagiri S. Discriminative training for large vocabulary speech recognition using minimum classification error. IEEE Trans. Audio, Speech Language Process. 15 1 (2007) 203-223
20. Misra, H., Bourlard, H., 2005. Spectral entropy feature in full-combination multi-stream for robust ASR. In: Proc. European Conf. Speech Communication and Technology, pp. 2633-2636.
21. Nadas A. A decision theoretic formulation of a training problem in speech recognition and a comparison of training by unconditional versus conditional maximum likelihood. IEEE Trans. Acoustics, Speech, Signal Process. 31 4 (1983) 814-817
22. Normandin, Y. Hidden, Markov Models, Maximum Mutual Information Estimation, and the Speech Recognition Problems. Ph.D. Dissertation, McGill University.
23. Ortmanns S., Ney H., and Aubert X. A word graph algorithm for large vocabulary continuous speech recognition. Comput. Speech Language 11 (1997) 43-72
24. Povey, D., 2004. Discriminative Training for Large Vocabulary Speech Recognition. Ph.D. Dissertation, Peterhouse, University of Cambridge.
25. Povey D., and Woodland P.C. Large scale discriminative training of acoustic models for speech recognition. Comput. Speech Language 16 (2002) 25-47
26. Povey, D., Woodland, P.C., 2002b. Minimum phone error and i-smoothing for improved discriminative training. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, pp. 105-108.
27. Saon G., and Padmanabhan M. Data-driven approach to designing compound words for continuous speech recognition. IEEE Trans. Speech Audio Process. 9 4 (2001) 327-332
28. Saon, G., Padmanabhan, M., Gopinath, R., Chen, S., 2000. Maximum likelihood discriminant feature spaces. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, vol. 2, pp. 1129-1132.
29. Stolcke, A., 2000. SRI language modeling toolkit. Version 1.3.3, 2000. .
30. Valchev, V., 1995. Discriminative Methods in HMM-based Speech Recognition. Ph.D. Dissertation, Peterhouse, University of Cambridge.
31. Vapnik V. The Nature of Statistical Learning Theory (1995), Springer-Verlag, New York
32. Wang H.M., Chen B., Kuo J.W., and Cheng S.S. MATBN: A Mandarin Chinese broadcast news corpus. Internat. J. Comput. Linguistic Chinese Language Process. 10 1 (2005) 219-235
33. Yu, D., Deng, L., 2007. Large-margin discriminative training of hidden markov models for speech recognition. In: Proc. IEEE Internat. Conf. Semantic Computing, pp. 429-438.
34. Yu, D., Deng, L., He, X., Acero A., 2007. Large-margin minimum classification error training for large-scale speech recognition tasks. In: Proc. IEEE Internat. Conf. Acoustics, Speech, Signal Processing, vol. 4, pp. 1137-1140.
35. Yu D., Deng L., He X., and Acero A. Large-margin minimum classification error training: A theoretical risk minimization perspective. Comput. Speech Language 22 4 (2008) 415-429