Home Artists Posts Import Register

Content

I can easily say that this will be the best models so far. Usually I had to choose for a better flow or less glitches, but this one seen to do both and better than the older ones.

Anime is almost complete, Life Action may take a few more days to be complete.

Here is a interpolation for the pokemon clip, at the start you can see some artifacts, but it probably will be fixed once the model finish training.

Both models should be ready by monday.

[Original]
https://streamable.com/022b6y

[Interpolated X2]
https://streamable.com/t3krwh

Once is ready and this new update is out, I already have plenty ideas to improve the model further, to bad that each training takes time.

Comments

Anonymous

Oh, hey. Thanks for working on this. I'm interested in this. Can totally see the smoothness difference between the two. Nice! As you probably know, there is a flickering issue with img2img when people try to create videos from frame-by-frame conversions of videos (after each video frame is extracted from said video and converted with img2img in Stable Diffusion). Do you think there is some way to deal with that problem with this app? :D

DAINAPP

flickering is a problem that is currently in all solutions for img2img I'm afraid. There been some solutions using ControlNet, but is far from perfect still. We will probably will need to wait for a really good txt2video model for that.