Fig. 6From: A multichannel learning-based approach for sound source separation in reverberant environmentsExample of the U-net for a two-source problemBack to article page