各种各样的卷积,卷积的设计

卷积

padding:

空洞卷积(atrous convolutions)又名扩张卷积(dilated convolutions)

transposed Convolutions)又名反卷积(deconvolution)

疑问:填充方式是怎样的?

Separable Convolution

Separable Convolution in Image Processing

$$ H = \begin{bmatrix}1 & 0 & -1 \ 2 & 0 & -2 \ 1 & 0 & -1 \end{bmatrix} = \begin{bmatrix}1 \ 2 \ -1 \end{bmatrix} * \begin{bmatrix}1 & 0 & -1 \end{bmatrix} $$

参考