A multi-band approach based on the probabilistic union model and frequency-filtering features for robust speech recognition.

Peter Jancovic, Ji Ming

Browse the full Interspeech paper archive.