3 ms·
Generally, the models seem to struggle with fine details that correlate with broader features. Each finger typically looks ok, it's the hand that's off... remot
by losteric 3y ago
Generally, the models seem to struggle with fine details that correlate with broader features. Each finger typically looks ok, it's the hand that's off... remote controls might have appropriately detailed buttons with a nonsense layout. I wonder if the training process can "weight" various visual features instead of optimizing loss equally over the whole image.
- IIAOPSW 3y agoYou're close, but there's something even more fundamentally wrong that goes overlooked. The noise model. Namely the use of gausssian random pixel flips. The spectrum of the noise distribution exactly gets in the way of the fine details before there's any chance to learn it. There's no chance for it to learn the layout of TV remote buttons because in the first iteration of the diffusion there's already noise of the same length scale as the buttons you want it to learn.