Skip to content

Dataset · tencent

Tencent 3D Object Generation Datasets

tencent/HY3D-Bench

Three 3D object collections with renders and points for text-to-3D and image-to-3D.

What's inside

HY3D-Bench includes three subsets: full objects with 252K+ watertight meshes with multi-view renders and sampled points; parts with 240K+ part-segmented objects; synthetic with 125K+ AI-generated objects across 1,252 categories. Tasks include text-to-3D and image-to-3D; metadata lists size as n>1T without units specified. The card mentions baselines and links to project, paper, and code.

What you can do with it

  • Train text-to-3D model on full 3D meshes
  • Fine-tune image-to-3D on synthetic objects
  • Evaluate part-segmented generation
Tasks
Text → 3d, Image → 3d
Size
over 1T rows
Downloads this month
36.9K
Likes
99

How to load it

from datasets import load_dataset

ds = load_dataset("tencent/HY3D-Bench")