Fig. 7From: Learning-based robust speaker counting and separation with the aid of spatial coherenceBlock diagram of the proposed speaker separation moduleBack to article page