[UPDATE] Using a larger point cloud model improves performance even more.

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

[UPDATE] Using a larger point cloud model improves performance even more. #5

Closed j0seo closed 1 year ago

j0seo commented 1 year ago

I just found that the default point cloud model was mistakenly typed as a smaller model (Point-E 40M). I have just pushed a commit that modifies this to use the 1B model, which produces much better results.

destlaver commented 1 year ago

can you post some of the results?