Abstract

Summary In recent years, there has been a trend towards training large vocabulary continuous speech recognition (LVCSR) systems on a large amount of found data. Found data is recorded from spontaneous speech without careful control of the recording acoustic conditions, for example, conversational telephone speech. Hence, it typically has greater variability in terms of speaker and acoustic conditions than specially collected data. Thus, in addition to the desired speech variability required to discriminate between words, it also includes various non-speech variabil-ities, for example, the change of speakers or acoustic environments. The standard approach to handle this type of data is to train hidden Markov models (HMMs) on the whole data set as if all data comes from a single acoustic condition. This is referred to as multi-style training, for exam-ple speaker-independent training. Effectively, the non-speech variabilities are ignored. Though good performance has been obtained with multi-style systems, these systems account for all variabilities. Improvement may be obtained if the two types of variabilities in the found data are modelled separately. Adaptive training has been proposed for this purpose. In contrast to multi-style training, a set of transforms is used to represent the non-speech variabilities. A canonical

Citations

8033	Maximum likelihood from incomplete data via the EM algorithm - Dempster, Laird, et al. - 1977
2635	Introduction to Statistical Pattern Recognition - Fukunaga - 1990
1908	Pattern Classification - Duda, Hart, et al. - 2001
1483	Fundamentals of Speech Recognition - Rabiner, Juang - 1993
1231	Statistical Decision Theory and Bayesian Analysis - Berger - 1985
1154	Error bounds for convolutional codes and an asymptotically optimum decoding algorithm - Viterbi - 1973
1104	Pattern Recognition and Neural Networks - Ripley - 1996
893	Monte Carlo Statistical Methods - Robert, Casella - 1999
752	Comparison of parametric representations for monosyllabic word recognition in continuously spoken sentences - Davis, Mermelstein - 1980
684	Optimal Statistical Decisions - DeGroot - 1970
591	Maximum Likelihood Linear Regression for Speaker Adaptation of Continuous Density - Leggetter, Woodland - 1995
489	Maximum a posteriori estimation for multivariate Gaussian mixture observations of Markov chains - Gauvain, Lee - 1994
458	A model for reasoning about persistence and causation - Dean, Kanazawa
406	Maximum Likelihood Linear Transformations For HMM-Based Speech Recognition - Gales - 1998
349	The Population Frequencies of Species and the Estimation of Population Parameters - Good - 1953
275	Variational algorithms for approximate Bayesian inference; PhD thesis: http://www.cse.buffalo.edu/faculty/mbeal/thesis - Beal - 1998
227	The zero-frequency problem: Estimating the probabilities of novel events in adaptive text compression - Witten, Bell - 1991
209	An Inequality With Applications to Statistical Estimation for Probabilistic Functions of a Markov Process and to a Model for Ecology - Baum, Egon - 1967
195	Some Statistical Issues in the Comparison of Speech Recognition Algorithms - Gillick, Cox
194	Accurate approximations for posterior moments and marginal densities - Tierney, Kadane - 1986
185	Effectiveness of linear prediction characteristics of the speech wave for automatic speaker identification and verification - Atal - 1974
181	Semi-tied covariance matrices for hidden Markov models - Gales - 1999
181	Continuous Speech Recognition by Statistical Methods - Jelinek - 1976
181	Tree-Based State Tying for High Accuracy Acoustic Modelling - Young, Odell, et al. - 1994
179	Minimum phone error and I-smoothing for improved discriminative training - Povey, Woodland - 2002
162	Maximum Mutual Information Estimation of Hidden Markov Model Parameters for Speech Recognition - Bahl, Brown, et al. - 1986
145	A compact model for speaker-adaptive training - Anastasakos - 1996
129	Minimum classification error rate methods for speech recognition - Juang, Chou, et al. - 1997
116	Discriminative training for large vocabulary speech recognition - Povey - 2003
109	Mean and variance adaptation within the mllr framework - Gales, Woodland - 1996
101	Confidence measures for large vocabulary continuous speech recognition - Wesse, Schluter, et al. - 2001
99	Large Scale Discriminative Training of Hidden Markov Models for Speech Recognition - Woodland, Povey
95	Rapid speaker adaptation in eigenvoice space - Kuhn, Junqua, et al.
89	Fast Speaker Adaptation Using Constrained Estimation of Gaussian Mixtures - Digalakis, Rtiscbev, et al. - 1995
83	The N-Best Algorithm: An Efficient and Exact Procedure for Finding the N Most Likely Sentence Hypotheses - Schwartz, Chow - 1990
82	Speaker normalization using efficient frequency warping procedures - Lee, Rose - 1996
81	Flexible speaker adaptation using maximum likelihood linear regression,” Eurospeech - Leggetter, Woodland - 1995
77	The Dragon System - an Overview - Baker - 1975
77	Investigation of silicon auditory models and generalization of linear discriminant analysis for improved speech recognition - Kumar - 1997
75	Structural maximum a posteriori linear regression for fast HMM adaptation - Siohan, Myrvoll, et al. - 2002
71	Large scale discriminative training for speech recognition - Woodland, Povey - 2000
70	An empirical Bayes approach to statistics - Robbins - 1955
64	The 1994 HTK large vocabulary speech recognition system - Woodland, Leggetter, et al. - 1995
63	ªMaximum Likelihood Estimation for Multivariate Mixture Observations of Markov Sources,º - Juang, Levinson, et al. - 1986
62	The generation and use of regression class trees for MLLR adaptation - Gales - 1996
57	Cluster Adaptive Training of Hidden Markov Models - Gales - 2000
57	Minimum Bayes-Risk Automatic Speech Recognition - Goel, Byrne - 2003
56	Large vocabulary continuous speech recognition using HTK - Woodland, Odell, et al. - 1994
51	RASTA-PLP speech analysis technique - Hermansky, Morgan, et al. - 1992
48	R.Schluter, “Using word probabilities as confidence measures - Wessel, Macherey - 1998