Open alexander-rakhlin opened 3 years ago
Is KernelGAN applicable for super resolution task without up-scaling, i.e. output = input size ? If yes, how such a regime can be implemented/turned on?
No. You are referring to de-blurring. Not SR. Similar but different
Is KernelGAN applicable for super resolution task without up-scaling, i.e. output = input size ? If yes, how such a regime can be implemented/turned on?