Skip to content

Dataset · tasl-lab

Unified 3D occupancy benchmark for autonomous driving

tasl-lab/uniocc

Unified 3D dataset and benchmark for predicting spatial occupancy in autonomous driving tasks.

What's inside

The set provides 3D data for tasks involving the prediction and forecasting of surrounding spatial occupancy around an autonomous vehicle. The dataset was created by the authors of the UniOcc project to bring disparate autonomous driving datasets into a unified format and semantics. The exact volume, languages, and file formats are not specified in the card.

What you can do with it

  • Evaluating and comparing 3D scene forecasting models for autopilots
  • Training environment perception systems from images (image-to-3D)
  • Unifying training pipelines across data from different driving datasets
Tasks
Image → 3d
Downloads this month
1.1M
Likes
46

How to load it

from datasets import load_dataset

ds = load_dataset("tasl-lab/uniocc")