This datasets contain two parts:
occupancies:
This part includes occupancies and point clouds. For specific usage, you can refer to Occupancy Networks.
rendered_images:
This part is a supplementary rendering dataset of Objaverse dataset. The rendering code is sourced from zero123, with the difference being the use of the Eevee renderer, and the camera positions are fixed at 12 locations on a sphere with a radius of 2. They are numbered from 12 to 23, corresponding to:
12: Front view
13:… See the full description on the dataset page:
https://huggingface.co/datasets/BAAI/Objaverse-MIX.
