Joint acoustic and spectral modeling for speech dereverberation using non-negative representations.

Nasser Mohammadiha, Paris Smaragdis, Simon Doclo

Browse the full ICASSP paper archive.