2 ms·
No, the paper means primarily the weight sharing in each kernel filter within each convolutional layer, and the stacking of these layers in deep networks.
by cs702 9y ago
No, the paper means primarily the weight sharing in each kernel filter within each convolutional layer, and the stacking of these layers in deep networks.