An EKF-based algorithm for learning statistical hidden dynamic model parameters for phonetic recognition

Author

Togneri, Roberto ; Deng, Li

Author_Institution

Univ. of Western Australia, WA, Australia

Volume

1

fYear

2001

fDate

2001

Firstpage

465

Abstract

Presents a parameter estimation algorithm based on the extended Kalman filter (EKF) for the statistical coarticulatory hidden dynamic model (HDM). We show how the EKF parameter estimation algorithm unifies and simplifies the estimation of both the state and parameter vectors. Experiments based on N-best rescoring demonstrate superior performance of the (context-independent) HDM over a triphone baseline HMM in the TIMIT phonetic recognition task. We also show that the HDM is capable of generating speech vectors close to those from the corresponding real data

Keywords

Kalman filters; filtering theory; hidden Markov models; nonlinear filters; parameter estimation; speech recognition; state estimation; N-best rescoring; TIMIT phonetic recognition task; extended Kalman filter; parameter estimation algorithm; phonetic recognition; speech vectors generation; statistical coarticulatory hidden dynamic model; triphone baseline HMM; Additive noise; Covariance matrix; Gaussian noise; Gaussian processes; Jacobian matrices; Multilayer perceptrons; Nonlinear equations; Parameter estimation; State estimation; Switches;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 2001. Proceedings. (ICASSP '01). 2001 IEEE International Conference on

Conference_Location

Salt Lake City, UT

ISSN

1520-6149

Print_ISBN

0-7803-7041-4

Type

conf

DOI

10.1109/ICASSP.2001.940868

Filename

940868