DescriptionThis code enables training and evaluation of a switching linear dynamic model for enhancing cepstral streams for automatic speech recognition, as described in our ICASSP 2004 paper, Noise Robust Speech Recognition with a Switching Linear Dynamic Model.
|