Sound source separation based on non-negative tensor factorization incorporating spatial cue as prior knowledge.

Yuki Mitsufuji, Axel Roebel

Browse the full ICASSP paper archive.