Skip to content

Dataset · hw-liang

Objaverse 3D and 4D object renders for generation

hw-liang/Diffusion4D

1 million to 10 million rendered videos and metadata of 3D/4D models for text-to-3D and image-to-3D tasks.

What's inside

A large dataset compiled from selected objects of the Objaverse-1.0 and Objaverse-XL databases for the Diffusion4D project. Inside are orbital videos of dynamic and static 3D objects, frontal-view monocular videos, and object metadata. The card specifies 3D modality, the English language, and a volume of 1 million to 10 million records.

What you can do with it

  • Training video diffusion models to generate 3D and 4D scenes from text or images
  • Training spatio-temporal consistency algorithms for 3D rendering
Tasks
Text → 3d, Image → 3d
Languages
English
Size
1M–10M rows
Downloads this month
1.1K
Likes
33

How to load it

from datasets import load_dataset

ds = load_dataset("hw-liang/Diffusion4D")