Maximum a posteriori adaptation for large scale HMM recognizers

Authors:
G. Zavaliagkos;R. Schwartz;J. McDonough
Affiliations:
BBN Syst. & Technol. Corp., Cambridge, MA, USA;-;-
Venue:
ICASSP '96 Proceedings of the Acoustics, Speech, and Signal Processing, 1996. on Conference Proceedings., 1996 IEEE International Conference - Volume 02
Year:
1996

Citing 0
Cited 5

Prior knowledge guided maximum expected likelihood based model selection and adaptation for nonnative speech recognition

Computer Speech and Language
Acoustic model adaptation based on pronunciation variability analysis for non-native speech recognition

Speech Communication
Automatic speech recognition and speech variability: A review

Speech Communication
On the effectiveness of robot-assisted language learning

ReCALL
Rapid speaker adaptation in latent speaker space with non-negative matrix factorization

Speech Communication

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present a framework for maximum a posteriori (MAP) adaptation of large scale HMM recognizers. First we review the standard MAP adaptation for Gaussian mixtures. We then show how MAP can be used to estimated transformations which are shared across many parameters. Finally, we combine both techniques: each of the HMM models is adapted based on an interpolation of MAP estimates obtained under varying degrees of sharing. We evaluate this algorithm for adaptation of a continuous density HMM with 96 K Gaussians and show that very satisfactory improvements can be achieved, especially for adaptation of non-native speakers of American English.