Comment by tmaly
1 day ago
it would be interesting if it could use a diffusion model to generate the bitmap then a different model to convert that bitmap to vector format. This could be an interesting way to reason about animations.
1 day ago
it would be interesting if it could use a diffusion model to generate the bitmap then a different model to convert that bitmap to vector format. This could be an interesting way to reason about animations.
No comments yet
Contribute on Hacker News ↗