Semi-supervised non-negative tensor factorisation of modulation spectrograms for monaural speech separation.

Tom Barker, Tuomas Virtanen

Browse the full IJCNN paper archive.