An Alphanet approach to optimising input transformations for continuous speech recognition.

John S. Bridle, L. Dodd

Browse the full ICASSP paper archive.