Fig. 2From: End-to-end speech emotion recognition using a novel context-stacking dilated convolution neural networka The dilated causal convolution block. b The local conditioningBack to article page