Abstract

This paper examines the application of linear transformations for speaker and environmental adaptation in an HMM-based speech recognition system. In particular, transformations that are trained in a maximum likelihood sense on adaptation data are investigated. Other than in the form of a simple bias, strict linear feature-space transformations are inappropriate in this case. Hence, only model-based linear transforms are considered. The paper compares the two possible forms of model-based transforms: (i) unconstrained, where any combination of mean and variance transform may be used, and (ii) constrained, which requires the variance transform to have the same form as the mean transform (sometimes referred to as feature-space transforms). Re-estimation formulae for all appropriate cases of transform are given. This includes a new and efficient "full" variance transform and the extension of the constrained model-space transform from the simple diagonal case to the full or block-diagonal case. The constrained and unconstrained transforms are evaluated in terms of computational cost, recognition time efficiency, and use for speaker adaptive training. The recognition performance of the two model-space transforms on a large vocabulary speech recognition task using incremental adaptation is investigated. In addition, initial experiments using the constrained model-space transform for speaker adaptive training are detailed. 1 The author is now at the IBM T.J. Watson Research Center, Yorktown Heights, NY 10598, USA 1

Citations

6232	Maximum likelihood from incomplete data via the em algorithm - Dempster, Laird, et al. - 1977
440	Maximum likelihood linear regression for speaker adaptation of continuous density hidden markov models - Leggetter, Woodland - 1995
139	Tree-based state tying for high accuracy acoustic modelling - Young, Odell, et al. - 1994
106	A compact model for speaker adaptive training - Anastasakos, McDonough, et al.
86	A maximum-likelihood approach to stochastic matching for robust speech recognition - Sankar, Lee - 1996
80	Mean and Variance Adaptation within the mllr Framework.” Computer Speech and Language 10 - Gales, Woodland - 1996
65	Fast speaker adaptation using constrained estimation of Gaussian mixtures - Digalakis, Rtischev, et al. - 1995
65	Model-based techniques for noise robust speech recognition - Gales - 1995
56	The 1994 HTK large vocabulary speech recognition system - Woodland, Leggetter, et al. - 1995
51	The Generation and Use of Regression Class Trees for mllr Adaptation - Gales - 1996
51	A One Pass Decoder Design For Large Vocabulary Recognition - Odell, Valtchev, et al. - 1994
44	Maximum a-posteriori estimation for multivariate Gaussian mixture observations of Markov chains - Gauvain, Lee - 1994
40	Probabilistic Optimum Filtering for Robust Speech Recognition - Neumeyer, Weintraub - 1994
30	Integrated models of signal and background with application to speaker identification - Rose, Hofstetter, et al. - 1994
29	Improved Acoustic Modelling for hmms using Linear Transformations - Leggetter - 1995
20	Experiments in speaker normalisation and adaptation for large vocabulary speech recognition - Pye, Woodland - 1997
16	Robust speech recognition based on stochastic matching - Sankar, Lee
14	Semi-tied full-covariance matrices for hidden Markov models - Gales - 1997
12	Speaker normalisation using efficient frequency warping procedures - Lee, Rose - 1996
12	Practical implementations of speaker adaptive training - Matsoukas, Schwartz, et al. - 1997
10	Unsupervised speaker adaptation by probabilistic spectrum fitting - Cox, Bridle - 1989
10	Unsupervised Speaker-Adaptation For Hybrid HMM-MLP Continuous Speech Recognition System - Neto, Martins, et al. - 1995
5	A Sankar, and V V Digalakis. A comparative study of speaker adaptation techniques - Neumeyer - 1995
4	Leggetter and P CWoodland. Maximum likelihood linear regression for speaker adaptation of continuous density HMMs. Computer Speech and Language - J - 1995