Psychoacoustic model compensation for robust continuous speech recognition in additive noise | IEEE Conference Publication | IEEE Xplore