Jesse R. Shue

Email  /  LinkedIn  /  Github  /  GScholar

profile photo
3D Neural Field Generation using Triplane Diffusion
J. Ryan Shue*, Eric Ryan Chan*, Ryan Po*, Zachary Ankner*, Jiajun Wu, Gordon Wetzstein
CVPR, 2023

By factoring a 3D dataset into axis-aligned triplanes, we can train state-of-the-art 2D diffusion models for 3D generation nearly out-of-the-box. We outperform prior art in 3D generation on three classes of ShapeNet.

The Canopy (2020-present)
Zenger3D (2021)
Zenger Writer (2021)

Thanks to Jon Barron for the website template.