Fig. 1From: Reverberant speech recognition exploiting clarity index estimationHistogram of C 50 values in the training setBack to article page