LASSO environment model combination for robust speech recognition

In this paper, we propose a novel acoustic model adaptation method
for noise robust speech recognition. Model combination is a com-
mon way to adapt acoustic models to a target test environment. For
example, the mean supervectors of the adapted model is obtained
as a linear combination of mean supervectors of many pre-trained
environment-dependent acoustic models. Usually, the combination
weights are estimated using a maximum likelihood (ML) criterion
and the weights are nonzero for all the mean supervectors. We pro-
pose to estimate the weights by using Lasso (least absolute shrink-
age and selection operator) which imposes an ?1 regularization term
in the weight estimation problem to shrink some weights to exactly
zero. Our study shows that Lasso usually shrinks to zero the weights
of those mean supervectors not relevant to the test environment. By
removing some nonrelevant supervectors, the obtained mean super-
vectors are found to be more robust against noise distortions. Ex-
perimental results on Aurora-2 task show that the Lasso-based mean
combination consistently outperforms ML-based combination.