Semi-supervised Monaural Singing Voice Separation with a Masking Network Trained on Synthetic Mixtures.

Michael Michelashvili, Sagie Benaim, Lior Wolf

Browse the full ICASSP paper archive.