Fig. 9From: Learning-based robust speaker counting and separation with the aid of spatial coherenceSettings for network training with different microphone array geometriesBack to article page