Abstract
It is well known that the introduction of acoustic background distortion into speech causes recognitionalgorithms to fail. In order to improve the environmental robustness of speech recognition inadverse conditions, a novel constrained-iterative feature-estimation algorithm, which was previouslyformulated for speech enhancement, is considered and shown to produce improved feature characterizationin a variety of actual noise conditions such as computer fan, large crowd, and voice...

This publication has 42 references indexed in Scilit: