Piecewise linear constraints for model space adaptation

Nguyen, Patrick;Rigazio, Luca;Junqua, Jean-Claude;Wellekens, Christian J
ICASSP 2002, 27th IEEE International Conference on Acoustics, Speech and Audio Processing, May 13-17, 2002, Orlando, USA

Setting linear constraints on HMM model space appears to be very effective for speaker adaptation. In doing so, we assume that model parameters are jointly Gaussian. While this
approach has proven reasonably successful, we question it accuracy in the case of very high dimensionality parameter spaces.
To address this problem, we employ a hierarchical piece-wise linear model. Gross speaker variations are modeled with a linear eigenspace, subsuming the joint Gaussian model, and finer residues are modeled using another eigenspace chosen depending on the location of the first values. We perform experiments on Wall Street Journal (WSJ) dictation task, and we observe a cumulative 1.3% WER improvement (11% relative) when using self-adaptation.

Digital Security
Eurecom Ref:
© 2002 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

PERMALINK : https://www.eurecom.fr/publication/751