Skip to content

Dataset · ziqima

Steer3D text-based 3D object editing data

ziqima/Steer3D-Data

A training set of images, metadata, and TRELLIS latent representations for text-based 3D model editing.

What's inside

The dataset was prepared by the authors of the research paper 'Feedforward 3D Editing via Text-Steerable Image-to-3D' for tasks involving element addition and removal, and texture modification. It contains source object images with unique identifiers, training metadata, and TRELLIS latent representations for source and target assets. The dataset size ranges from 10K to 100K records.

What you can do with it

  • Training image-to-3D generation models
  • Training neural networks for text-driven 3D scene and texture editing
Tasks
Image → 3d
Size
10K–100K rows
Downloads this month
24.2K
Likes
0

How to load it

from datasets import load_dataset

ds = load_dataset("ziqima/Steer3D-Data")