Well with video in particular I have to cut short the GAN part of training that I do much longer for images that makes them more vibrant. Otherwise glitches start getting introduced that are particularly noticeable in video. It’s an unfortunate tradeoff.
-
-
That being said, both images and video are trained purely based on RGB values rather than something like HSV etc. I’ve been thinking about this lately actually. It seems it may make a difference to move away from RGB as it may be promoting averaging.
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.