In case you are wondering about this, you don't need GANs because there is a ground truth. You are just converting noisy images into denoised images. There is literature on "image restoration" showing hybrid L1/perceptual loss do best (http://arxiv.org/pdf/1511.08861 ).https://twitter.com/citnaj/status/1272759311639310336 …
-
-
It would be great if such general insights were listed somewhere in a curated fashion. I'm left to assume many AI practitioners are regularly reinventing the wheel. The pace of change is so fast its a challenge to keep up and not use out dated tech from last year, if it seems ok.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.