Fig. 4From: A multichannel learning-based approach for sound source separation in reverberant environmentsThe structure of the proposed network based on two training stagesBack to article page