In this paper, a framework for maximum a posteriori (MAP) estimation of hidden Markov models (HMM) is presented. Three key issues of MAP estimation, namely, the choice of prior distribution family, the specification of the parameters of prior densities, and the evaluation of the MAP estimates, are addressed. Using HMM's with Gaussian mixture state observation densities as an example, it is assumed that the prior densities for the HMM parameters can be adequately represented as a product of Dirichlet and normal-Wishart densities. The classical maximum likelihood estimation algorithms, namely, the forward-backward algorithm and the segmental k-means algorithm, are expanded, and MAP estimation formulas are developed. Prior density estimation issues are discussed for two classes of applications-parameter smoothing and model adaptation-and some experimental results are given illustrating the practical interest of this approach. Because of its adaptive nature, Bayesian learning is shown to serve as a unified approach for a wide range of speech recognition applications.
%0 Journal Article
%1 Gauvain1994
%A Gauvain, Jean-Luc
%A Lee, Chin-Hui
%D 1994
%J IEEE Transactions on Speech and Audio Processing
%K Bayes Markov a adaptation;normal-Wishart algorithm;hidden algorithm;speech algorithms;model data densities;Adaptation densities;segmental density;Gaussian density;parameter estimation estimation;Maximum estimation;Parameter estimation;Robustness;Smoothing estimation;forward-backward estimation;maximum estimation;parameter estimation;speech k-means learning;Dirichlet likelihood methods;Hidden methods;Speech methods;hidden mixture;HMM model;Bayesian models;Maximum models;maximum observation parameters;MAP posteriori processes;Bayesian recognition;Training recognition;state recognition;stochastic smoothing;prior
%N 2
%P 291-298
%R 10.1109/89.279278
%T Maximum a posteriori estimation for multivariate Gaussian mixture observations of Markov chains
%V 2
%X In this paper, a framework for maximum a posteriori (MAP) estimation of hidden Markov models (HMM) is presented. Three key issues of MAP estimation, namely, the choice of prior distribution family, the specification of the parameters of prior densities, and the evaluation of the MAP estimates, are addressed. Using HMM's with Gaussian mixture state observation densities as an example, it is assumed that the prior densities for the HMM parameters can be adequately represented as a product of Dirichlet and normal-Wishart densities. The classical maximum likelihood estimation algorithms, namely, the forward-backward algorithm and the segmental k-means algorithm, are expanded, and MAP estimation formulas are developed. Prior density estimation issues are discussed for two classes of applications-parameter smoothing and model adaptation-and some experimental results are given illustrating the practical interest of this approach. Because of its adaptive nature, Bayesian learning is shown to serve as a unified approach for a wide range of speech recognition applications.
@article{Gauvain1994,
abstract = {In this paper, a framework for maximum a posteriori (MAP) estimation of hidden Markov models (HMM) is presented. Three key issues of MAP estimation, namely, the choice of prior distribution family, the specification of the parameters of prior densities, and the evaluation of the MAP estimates, are addressed. Using HMM's with Gaussian mixture state observation densities as an example, it is assumed that the prior densities for the HMM parameters can be adequately represented as a product of Dirichlet and normal-Wishart densities. The classical maximum likelihood estimation algorithms, namely, the forward-backward algorithm and the segmental k-means algorithm, are expanded, and MAP estimation formulas are developed. Prior density estimation issues are discussed for two classes of applications-parameter smoothing and model adaptation-and some experimental results are given illustrating the practical interest of this approach. Because of its adaptive nature, Bayesian learning is shown to serve as a unified approach for a wide range of speech recognition applications.},
added-at = {2021-02-01T10:51:23.000+0100},
author = {Gauvain, Jean-Luc and Lee, Chin-Hui},
biburl = {https://www.bibsonomy.org/bibtex/246dd42b26b083b69b2924a4d7b7d27e4/m-toman},
doi = {10.1109/89.279278},
file = {:pdfs/gauvain_ieeetrans_1994.pdf:PDF},
interhash = {d91fb0ec5d0c498ea576b0b32b99e49a},
intrahash = {46dd42b26b083b69b2924a4d7b7d27e4},
issn = {1063-6676},
journal = {IEEE Transactions on Speech and Audio Processing},
keywords = {Bayes Markov a adaptation;normal-Wishart algorithm;hidden algorithm;speech algorithms;model data densities;Adaptation densities;segmental density;Gaussian density;parameter estimation estimation;Maximum estimation;Parameter estimation;Robustness;Smoothing estimation;forward-backward estimation;maximum estimation;parameter estimation;speech k-means learning;Dirichlet likelihood methods;Hidden methods;Speech methods;hidden mixture;HMM model;Bayesian models;Maximum models;maximum observation parameters;MAP posteriori processes;Bayesian recognition;Training recognition;state recognition;stochastic smoothing;prior},
month = apr,
number = 2,
owner = {schabus},
pages = {291-298},
timestamp = {2021-02-01T10:51:23.000+0100},
title = {Maximum a posteriori estimation for multivariate Gaussian mixture observations of Markov chains},
volume = 2,
year = 1994
}