News New AI turns image to 3D Model (Post some)

[veryrare]

MIDI 3D - a Hugging Face Space by VAST-AI

Image to Compositional 3D Scene Generation

huggingface.co

GitHub - VAST-AI-Research/MIDI-3D: Official impl. of "MIDI: Multi-Instance Diffusion for Single Image to 3D Scene Generation" (CVPR 2025)

Official impl. of "MIDI: Multi-Instance Diffusion for Single Image to 3D Scene Generation" (CVPR 2025) - VAST-AI-Research/MIDI-3D

github.com

Here's how to use it:

Don't use super complicated shit It's gonna time out, Also I think you need to make an account so yeah do that

@wereq @DarkStar @Sergeant Kelly @Stupid Clown

 

[Paladin]

damn

 

[Sergeant Kelly]

Huh, I'm going to try it out although I suspect right now it's like how it used to be with text-to-image generation couple of years ago - more of a curiosity than something useful until it gets developed more.
Could be good for making simple assets though

 

[veryrare]

damn

View attachment 1409877

Nigga, watch the tutorial
You're supposed to draw boxes around the shit

Huh, I'm going to try it out although I suspect right now it's like how it used to be with text-to-image generation couple of years ago - more of a curiosity than something useful until it gets developed more.
Could be good for making simple assets though

My thoughts exactly

 

[Paladin]

Nigga, watch the tutorial
You're supposed to draw boxes around the shit

Dnw but it probably wouldn't work anywhere

 

[DarkStar]

very interesting, noted