Non-linear maximum likelihood feature transformation for speech recognition.

Mohamed Kamal Omar, Mark Hasegawa-Johnson

Browse the full Interspeech paper archive.