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

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-...

Full description

Saved in:
Bibliographic Details
Published in2001 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.01CH37221) Vol. 1; pp. 465 - 468 vol.1
Main Authors Togneri, R., Li Deng
Format Conference Proceeding
LanguageEnglish
Published IEEE 2001
Subjects
Additive noise
Covariance matrix
Gaussian noise
Gaussian processes
Jacobian matrices
Multilayer perceptrons
Nonlinear equations
Parameter estimation
State estimation
Switches
Online AccessGet full text
ISBN0780370414
9780780370418
ISSN1520-6149
DOI10.1109/ICASSP.2001.940868

Cover

More Information
Summary: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.
ISBN:0780370414
9780780370418
ISSN:1520-6149
DOI:10.1109/ICASSP.2001.940868