Unsupervised training of acoustic models for large vocabulary continuous speech recognition

IEEE Transactions on Speech and Audio Processing

Volumn 13, Issue 1, 2005, Pages 23-31

  Wessel, Frank ^a   Ney, Hermann ^b  

^a Vodafone D2 GmbH   (Germany)

^b RWTH AACHEN UNIVERSITY   (Germany)

Author keywords

Confidence measures; Speech recognition; Unsupervised training; Word graph; Word posterior probabilities

Indexed keywords

ACOUSTICS; ERROR ANALYSIS; ITERATIVE METHODS; MARKOV PROCESSES; MATHEMATICAL MODELS;

CONFIDENCE MEASURES; UNSUPERVISED TRAINING; WORD GRAPHS;

SPEECH RECOGNITION;

EID: 11144239919     PISSN: 10636676     EISSN: None     Source Type: Journal    
DOI: 10.1109/TSA.2004.838537     Document Type: Article

References (18)

1. G. Zavaliagkos and T. Colthurst, "Utilizing untranscribed training data to improve performance," in Proc. Broadcast News Transcription and Understanding Workshop, San Mateo, CA, Feb. 1998, pp. 301-305.
2. M. Siu, H. Gish, and F. Richardson, "Improved estimation, evaluation, and applications of confidence measures for speech recognition," in Proc. ISCA Eur. Conf. Speech Communication Technology, Rhodes, Greece, Sep. 1997, pp. 831-834.
3. T. Kemp and A. Waibel, "Unsupervised training of a speech recognizer: Recent experiments," in Proc. ISCA Eur. Conf. Speech Communication Technology, Budapest, Hungary, Sep. 1999, pp. 2725-2728.
4. L. Lamel, J. L. Gauvain, and G. Adda, "Lightly supervised acoustic model training," in Proc. Automatic Speech Recognition Workshop, Sep. 2000, pp. 150-154.
5. _, "Investigating lightly supervised acoustic model training," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing, Salt Lake City, UT, May 2001, pp. 477-480.
6. _, "Lightly supervised and unsupervised acoustic model training," Comput., Speech, Lang., vol. 16, pp. 115-129, 2002.
7. F. Wessel and H. Ney, "Unsupervised training of acoustic models for large vocabulary continuous speech recognition," in Proc. Automatic Speech Recognition Workshop, Madonna di Campiglio, Trento, Italy, Dec. 2001.
8. H. Ney, "Acoustic modeling of phoneme units for continuous speech recognition," in Proc. Signal Processing V: Theories and Applications, 5th Eur. Signal Processing Conference, L. Torres, E. Masgrau, and M. A. Lagunas, Eds, Barcelona, Spain, 1990, pp. 65-72.
9. R. Schwartz, Y.-L. Chow, O. Kimball, S. Roukos, U. Krasner, and J. Makhoul, "Context-dependent modeling for acoustic-phonetic recognition of continuous speech," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing, Tampa, FL, Mar. 1985, pp. 1205-1208.
10. H. Ney and A. Noll, "Phoneme modeling using continuous mixture densities," in Proc. IEEE Int. Conf, Acoustics, Speech, Signal Processing, New York, Apr. 1988, pp. 437-440.
11. F. Wessel, R. Schlüter, K. Macherey, and H. Ney, "Confidence measures for large vocabulary continuous speech recognition," IEEE Trans. Speech Audio Process., vol. 9, no. 3, pp. 288-298, Mar. 2001.
12. F. Wessel. (2002, Aug.) Word posterior probabilities for large vocabulary continuous speech recognition. Univ. Techno. (RWTH) Aachen, Germany. [Online], Available: http://sylvester.bth.rwth-aachen.de/ dissertationen/2002/121/02_121. pdf
13. P. C. Woodland, C. J. Legetter, J. J. Odell, V. Valtchev, and S. J. Young, "The 1994 HTK large vocabulary speech recognition system," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing, Detroit, MI, May 1995, pp. 573-576.
14. S. Ortmanns, H. Ney, and X. Aubert, "A word graph algorithm for large vocabulary continuous speech recognition," Comput., Speech, Lang., vol. 11.no. 1, pp. 43-72, Jan. 1997.
15. T. Kemp, "ILKD, Ein automatisches Indexierungssystem für Fernsehnachrichtensendungen," Ph.D. thesis, Tech. Univ. Karlsruhe, Karlsruhe, Germany, 1999.
16. J. S. Garofolo, J. G. Fiscus, and W. M. Fisher, "Design and preparation of the 1996 HUB-4 broadcast news benchmark test corpora," in Proc. Speech Recognition Workshop, San Francisco, CA, Feb. 1997, pp. 15-21.
17. D. S. Pallett, J. G. Fiscus, J. S. Garofolo, A. Martin, and M. A. Przybocki, "1998 broadcast news benchmark test results: English and nonenglish word error rate performance measures," in Proc. DARPA Broadcast News Workshop, Herndon, VA, Feb. 1999, pp. 5-12.
18. A. Sixtus and H. Ney, "From within-word model search to across-word model search in large vocabulary continuous speech recognition," Computer, Speech, Lang., 2001, submitted for publication.