The new "all in one" photo restoration model I'm working on is still early in training, but it's already working quite well to my surprise. Samples:pic.twitter.com/S65pyoqcku
Show this thread
-
Does it perform better restoring colorized pictures or colorizing restored pictures?
I haven't gotten that far yet honestly. The intention is that you restore first then colorize because that just makes more sense but I won't know for sure until it's more thoroughly tested.
-
-
I had read a bit last year about networks doing two tasks together (I don't remember how it's called), and how it benefited both tasks. Is that something that would make sense here?
-
It's actually inspired by that and similar findings :) Like..basically I'm getting much better generalization by doing this swiss army knife approach. I don't want to give everything away but I can tell you I didn't explicitly train for the strip patterns in that one b/w image
End of conversation
New conversation -
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.