search

KU-CVLAB / 3DFuse

Official implementation of "Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation"
711 stars 42 forks source link

Custom SD-based models #7

Closed ReRadioActivity closed 1 year ago

ReRadioActivity commented 1 year ago

Is it possible to use any other custom SD-based model with this repo and get a decent result? Or only the ones that are presented in the Gradio demo work properly?

j0seo commented 1 year ago

We used SD v1.5 in our framework. It's difficult to answer whether other custom SDs can be used since we haven't tried them. However, if the sparse depth injector is newly trained with custom SDs, we expect that it will also perform well.

If the model you are referring to is not a diffusion model but a model used for representation like NeRF (e.g., DMTet), we also expect it to work well.