I need to implement a segmentation-aware conv operation, which means during conv, the weights of the filter is multiplied by a local segmentation-aware weights.
As the weights are different in different location, i need to unfold the image conv to act like matrix multiplcation.
you could either use
pytorch or im2col/col2im from https://github.com/szagoruyko/pyinn
will unfold support dilated spacing in future? with which the dilated convolution can be implemented too.