A Hands-On Comparison of DNNs for Dialog Separation Using Transfer Learning from Music Source Separation.

Martin Strauss, Jouni Paulus, Matteo Torcoli, Bernd Edler

Browse the full Interspeech paper archive.