Construction of model-space constraints

Abstract

HMM systems exhibit a large amount of redundancy. To this end, a technique called eigenvoices was found to be very effective for speaker adaptation. The correlation between HMM parameters is exploited via a linear constraint called eigenspace. This constraint is obtained through a PCA of the training speakers. We show how PCA can be linked to the maximum-likelihood criterion. Then, we extend the method to LDA transformations and piecewise linear constraints. On the Wall Street Journal (WSJ) dictation task, we obtain 1.7% WER improvement (15% relative) when using self-adaptation.