Abstract
This paper deals with the problem of efficient speaker adaptation in large vocabulary continuous speech recognition (LVCSR) systems. The main goal is to adapt acoustic models of speech and to increase the recognition accuracy of these systems in tasks, where only one user is expected (e.g. voice dictation) or where the speaking person can be identified automatically (e.g. broadcast news transcription). For this purpose, we propose several modifications of the well known MLLR (Maximum Likelihood Linear Regression) method and we combine them with the MAP (Maximum A Posteriori) method. The results from a series of experiments show that the error rate of our 300K-word Czech recogniser can be reduced by about 9.9 % when only 30 seconds of supervised data are used for adaptation or by about 9.6 % when unsupervised adaptation on the same data is performed.
This work was supported by the Czech Grant Agency in project no. 102/05/0278.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Woodland, P.C.: Speaker Adaptation: Techniques and Challenges. In: Proc. IEEE Workshop on Automatic Speech Recognition and Understanding, Keystone (1999)
Gauvain, J.L., Lee, C.H.: Maximum A Posteriori Estimation for Multivariate Gaussian Mixture Observations of Markov Chains. IEEE Trans. SAP 2, 291–298 (1994)
Leggetter, C.J., Woodland, P.C.: Flexible Speaker Adaptation Using Maximum Likelihood Linear Regression. In: Proc. ARPA Spoken Language Technology Workshop, pp. 104–109. Morgan Kaufmann, San Francisco (1995)
Nouza, J., Nejedlova, D., Zdansky, J., Kolorenc, J.: Very Large Vocabulary Speech Recognition System for Automatic Transcription of Czech Broadcast Programs. In: Proc. of Int. Conference on Spoken Language Processing (ISCLP 2004), Jeju (October 2004)
Huang, X.D., Acero, A., Hon, H.W.: Spoken Language Processing. Prentice-Hall, Englewood Cliffs (2001)
Gales, M.J.F., Woodland, P.C.: Mean and Variance Adaptation Within the MLLR Framework. Computer Speech and Language 10, 249–264 (1996)
Nouza, J., Psutka, J., Uhlir, J.: Phonetic Alphabet for Speech Recognition of Czech. Radioengineering 6(4), 16–20 (1997)
Zelezny, M.: Speaker adaptation in continuous speech recognition system of Czech. PhD thesis (in Czech). Z ČU of Plzeň (2001)
Chesta, C., Siohan, O., Lee, C.H.: Maximum a posteriori linear regression for hidden Markov model adaptation. In: Proceedings of European Conference on Speech Communication and Technology, Budapest, Hungary, vol. 1, pp. 211–214 (1999)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cerva, P., Nouza, J. (2005). Supervised and Unsupervised Speaker Adaptation in Large Vocabulary Continuous Speech Recognition of Czech. In: Matoušek, V., Mautner, P., Pavelka, T. (eds) Text, Speech and Dialogue. TSD 2005. Lecture Notes in Computer Science(), vol 3658. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11551874_26
Download citation
DOI: https://doi.org/10.1007/11551874_26
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-28789-6
Online ISBN: 978-3-540-31817-0
eBook Packages: Computer ScienceComputer Science (R0)