Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

A Robust Training Strategy Against Extraneous Acoustic Variations for Spontaneous Speech Recognition

Hui Jiang (1,2), Li Deng (1,3)

(1) Department of Electrical and Computer Engineering, University of Waterloo, Canada
(2) Currently Multimedia Communications Research Lab, Bell Labs, Murray Hill, NJ, USA
(3) Currently Microsoft Research, Redmond, WA, USA

In the paper, we propose a robust training strategy to deal with extraneous acoustic variations for conversational speech recognition. This strategy generalizes speaker adaptive training, where HMM parameter transformations are used to normalize the extraneous variations in the training data according to a set of pre-defined conditions. Then a compact model and the associated prior p.d.f.ís of transformation parameters are estimated using the maximum likelihood criterion. In the testing phase, the compact model and the prior p.d.f.ís are used to search for the unknown word sequence based on Bayesian Prediction Classification. The proposed strategy is evaluated in a Switchboard task to deal with pronunciation variations in spontaneous speech recognition. Preliminary results show moderate word error rate reduction over a well-trained baseline system under identical experimental conditions.


Full Paper

Bibliographic reference.  Jiang, Hui / Deng, Li (2000): "A robust training strategy against extraneous acoustic variations for spontaneous speech recognition", In ICSLP-2000, vol.4, 161-164.