ML estimation of a stochastic linear system with the EM algorithm and its application to speech recognition

Author

Digalakis, V. ; Rohlicek, J.R. ; Ostendorf, M.

Author_Institution

SRI Int., Menlo Park, CA, USA

Volume

1

Issue

4

fYear

1993

fDate

10/1/1993 12:00:00 AM

Firstpage

431

Lastpage

442

Abstract

A nontraditional approach to the problem of estimating the parameters of a stochastic linear system is presented. The method is based on the expectation-maximization algorithm and can be considered as the continuous analog of the Baum-Welch estimation algorithm for hidden Markov models. The algorithm is used for training the parameters of a dynamical system model that is proposed for better representing the spectral dynamics of speech for recognition. It is assumed that the observed feature vectors of a phone segment are the output of a stochastic linear dynamical system, and it is shown how the evolution of the dynamics as a function of the segment length can be modeled using alternative assumptions. A phoneme classification task using the TIMIT database demonstrates that the approach is the first effective use of an explicit model for statistical dependence between frames of speech

Keywords

hidden Markov models; linear systems; maximum likelihood estimation; parameter estimation; speech recognition; stochastic processes; EM algorithm; ML estimation; TIMIT database; dynamical system model; expectation-maximization algorithm; feature vectors; hidden Markov models; parameter estimation; phone segment; phoneme classification; segment length; spectral dynamics; speech frames; speech recognition; stochastic linear dynamical system; Hidden Markov models; Kalman filters; Linear systems; Maximum likelihood estimation; Parameter estimation; Speech recognition; State estimation; Stochastic systems; Time varying systems; Vectors;

fLanguage

English

Journal_Title

Speech and Audio Processing, IEEE Transactions on

Publisher

ieee

ISSN

1063-6676

Type

jour

DOI

10.1109/89.242489

Filename

242489