9th Annual Conference of the International Speech Communication Association

Brisbane, Australia
September 22-26, 2008

Parameter Clustering and Sharing in Variable-Parameter HMMs for Noise Robust Speech Recognition

Dong Yu (1), Li Deng (1), Yifan Gong (2), Alex Acero (1)

(1) Microsoft Research, USA; (2) Microsoft Corporation, USA

Recently we proposed a cubic-spline-based variable-parameter hidden Markov model (CS-VPHMM) whose mean and variance parameters vary according to some cubic spline functions of additional environment-dependent parameters. We have shown good properties of the CS-VPHMM and demonstrated on the Aurora-3 corpus that MCE-trained CS-VPHMM greatly outperforms the MCE-trained conventional HMM at the cost of increased total number of model parameters. In this paper, we propose to share spline functions across different Gaussian mixture components to reduce the total number of model parameters and develop a clustering algorithm to do so. We demonstrate the effectiveness of our parameter clustering and sharing algorithm for the CS-VPHMM on Aurora-3 corpus and show that proper parameter sharing can reduce the number of parameters from 4 times of that used in the conventional HMM to 1.13 times and still get 18% relative WER reduction over the MCE trained conventional HMM under the well-matched condition. Effective parameter sharing makes the CS-VPHMM an attractive model for noise robustness.

Full Paper

Bibliographic reference.  Yu, Dong / Deng, Li / Gong, Yifan / Acero, Alex (2008): "Parameter clustering and sharing in variable-parameter HMMs for noise robust speech recognition", In INTERSPEECH-2008, 1253-1256.