Comment by DavidSJ
3 years ago
It may not be expensive if it’s being approximately done by a neural network that implicitly learned to do it when sharpening images.
3 years ago
It may not be expensive if it’s being approximately done by a neural network that implicitly learned to do it when sharpening images.
this feels more like wild speculation than a plausible explanation
Seems at least as likely that the neural network would learn a generally-applicable approximate Gaussian deconvolution as a specifically-applicable Moon upscale.