Adversarial Semi-Supervised Audio Source Separation Applied to Singing Voice Extraction.

Daniel Stoller, Sebastian Ewert, Simon Dixon

Browse the full ICASSP paper archive.