Generation of robust phonetic set and decision tree for Mandarin using chi-square testing

Speech Communication

Volumn 38, Issue 3-4, 2002, Pages 349-364

  Chen, Yeou Jiunn ^a   Wu, Chung Hsien ^a   Chiu, Yu Hsien ^a   Liao, Hsiang Chuan ^a  

^a NATIONAL CHENG KUNG UNIVERSITY   (Taiwan)

Author keywords

Confusion characteristic; Decision tree; Mandarin speech recognition; Phonetic representation

Indexed keywords

ACOUSTIC WAVES; SPEECH RECOGNITION; SPEECH SYNTHESIS; STATISTICAL METHODS; TREES (MATHEMATICS);

ACOUSTIC MODELING; CHI-SQUARE TESTING;

SPEECH ANALYSIS;

EID: 0036836706     PISSN: 01676393     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0167-6393(01)00076-0     Document Type: Article

References (23)

1. Agresti A. Categorical Data Analysis. 1990;Wiley, New York.
2. Bahl L.R., de Souza P.V., Gopalakrishnan P.S., Nahamoo D., Picheny M.A. Decision trees for phonological rules in continuous speech. Proc. ICASSP91, Toronto. 1991;185-188.
3. Bahl L.R., de Souza P.V., Gopalakrishnan P.S., Nahamoo D., Picheny M.A. Robust methods for using context-dependent features and models in a continuous speech recognizer. Proc. ICASSP94, Adelaide. 1997;533-536.
4. Boulianne G., Kenny P. Optimal tying of HMM mixture densities using decision trees. Proc. ICSLP96, Philadelphia. 1996;350-353.
5. Chen S.H., Liao Y.F., Chiang S.M., Chang S. An RNN-based pre-classification method for fast continuous Mandarin speech recognition. IEEE Trans. Speech Audio Process. 6(1):1998;86-90.
6. Cochran W.G. Some methods for strengthening of common tests. J. Internat. Biometric Soc. 1954;417-451.
7. Duchateau J., Demuynck K., Van Compernolle D. A novel node splitting criteria in decision tree construction for semi-continuous HMMs. Proc. Eurospeech97. Vol. 3:1997;1183-1186.
8. Huo Q., Ma B. Irrelevant variability normalization in learning HMM state tying from data based on phonetic decision-tree. Proc. ICASSP99. Vol. 2:1999;577-580.
9. Hwang, M.Y., Huang, X., 1998. Dynamically configurable acoustic models for speech recognition. In: Proc. ICASSP98, Seattle, pp. 669-672.
10. Hwang M.Y., Huang X., Alleva F. Predicting unseen triphones with senones. Proc. ICASSP93, Minneapolis. 1993;311-314.
11. Krzanowski W.J. Principles of Multivariate Analysis. 1988;Oxford University Press, New York.
12. Kuhn R., Lazarides A., Normandin Y., Brousseau J. Improved decision trees for phonetic modeling. Proc. ICASSP95, Detroit. 1995;552-555.
13. Lang, B.I., 1998. Acoustic models for continuous Mandarin speech recognition. Master Thesis, University of Taiwan, Taiwan, ROC.
14. Lyu R.Y., Hong I.C., Shen J.L., Lee M.Y., Lee L.S. Isolated Mandarin based-syllable recognition based upon the segmental probability model. IEEE Trans. Speech Audio Process. 6(3):1998;293-299.
15. Mathews, R.H., 1975. Mathews' Chinese-English Dictionary, Caves, 13th printing.
16. Odell, J.J., 1995. The use of context in large vocabulary speech recognition. Ph.D. Thesis, University of Cambridge, UK.
17. Rabiner L., Juang B.H. Fundamentals of Speech Recognition. 1993;Prentice-Hall, Englewood Cliffs, NJ.
18. Seide F., Wang N.J.C. Phonetic modeling in the Philips Chinese continuous-speech recognition system. Proc. ISCSLP'98, Singapore. 1998;54-59.
19. Seide F., Huang M., Huang H., Che C.W., Wang N.J.C. SAMPA-C - A phonetic representation of Chinese for speech recognition. Proc. ORIENTAL COCOSDA'99. 1999;105-108 Academia Sinica, Taipei, Taiwan.
20. Wells, J., 1997. EAGGLES Handbook on Spoken Language Systems (DRAFT) - SAMPA computer readable phonetic alphabet. www.phon.ucl.ac.uk/home/sampa/home.htm.
21. Willett D., Neukirchen C., Rottland J., Rigoll G. Refining tree-based state clustering by means of formal concept analysis, balanced decision trees and automatically generated model sets. Proc. ICASSP99. Vol. 2:1999;565-568.
22. Wu C., Reichl W. Decision tree state tying based on penalized Bayesian information criteria. Proc. ICASSP99. Vol. 1:1999;345-348.
23. Wu C.H., Chen Y.J., Yan G.L. Integration of phonetic and prosodic information for robust utterance verification. IEE Proc. Vision Image and Signal Process. 147:2000;55-61.