Fig. 3
From: Frequency-dependent auto-pooling function for weakly supervised sound event detection
Block diagram of the depthwise separable convolution with dilated filter in DDC-block
Fig. 3 | EURASIP Journal on Audio, Speech, and Music Processing