sudo-ai-3d / one2345plus Goto Github PK

Hello, very nice work! I was exploring a bit with the demo and was wondering if there is potential to apply this to character/facial 3D generation? It works pretty well on objects in a one-shot setting (especially when they have a good axial symmetry), but it is much more challenging on generation of humans/characters.

Do you think, for instance, that in a few shot setting the diffusion model that generates the novel 2D views can be skipped by providing multiple views at the start? Or, would it be doable to adopt a fine-tuning approach on a dataset specifically focused on 3D characters rendering? Thanks a lot!

Will you release the code?

Hey guys, wonderful work.

I was wondering, if I can inject e.g. 4 real images, I shot, from around an object like 45° each and use ur 3D object gen for interpolating in between the images e.g. 4 more missing images, is this something u are working on?

Or is it possible to inject 4 images or more into ur pipeline and switching off the zero123 generation but using the given images of which I would add an angle estimator to have the perspective as prior for ur 3d mesh generator?

Looking forward.

Hi!

I tried the demo on https://www.sudo.ai/3dgen and it's both fast and good quality, awesome work!
Was wondering if there'll be a hosted API endpoint service for usage?

When will the supplementary material be released? I am very interested in how you use the 2D features extracted by clip in the second stage of diffusion. I also use a similar method in the work I am doing now.

Hi 👋

First of all great work you did! I love the simplicity of the solution.

Do you plan to share the weights of the 3 diffusion networks?

Thank you for your work. I would like to know how you generated the video in the screenshot through AI? Do you need to work with other 3D software?

Do you have any plan to release the code of the model?

I'm interested in the code. when we can get the code release.