Ultrasound-coupled semi-supervised nonnegative matrix factorisation for speech enhancement.

Tom Barker, Tuomas Virtanen, Olivier Delhomme

Browse the full ICASSP paper archive.