3 ms·
VFusion3D: Learning Scalable 3D Generative Models from Video Diffusion Models
- MrTrvp 2y agohttps://arxiv.org/abs/2403.12034 https://arxiv.org/abs/2403.12034
- billconan 2y agoI tried a few samples, I feel that its quality is not as good as stable fast 3d.
- vessenes 2y agoInteresting that they compare to totally different models than Stable Fast 3D’s paper. I’m not clear on relative size of VFusion3D vs Stable Fast 3D, but I do think the training idea is good and novel — getting relatively good quality out of movies is a much easier ask in terms of collecting training data than getting 3D model renderings.
- corysama 2y agoI'm guessing this paper is more about "It's neat that this works at all." rather than trying to improve on the state of the art.
- iTokio 2y agoAnd stable fast 3D main point wasn’t quality but speed.