Multichannel Speech Separation with Recurrent Neural Networks from High-Order Ambisonics Recordings.

Laurline Perotin, Romain Serizel, Emmanuel Vincent, Alexandre Gurin

Browse the full ICASSP paper archive.