Skip to content

Dataset · CM-EVS-Dataset

Sparse RGB-D panorama evaluation dataset for 3D scenes

CM-EVS-Dataset/cmevs-erp-eval

Panoramic RGB-D data with camera poses for evaluating geometric coverage quality of 3D scenes.

What's inside

The CM-EVS dataset contains 10 to 100 thousand records and comprises a selection of panoramic RGB-D data with camera pose information. The dataset is designed to provide maximum geometric coverage of a 3D scene with a minimal number of equirectangular projection (ERP) frames. This release part is intended for model evaluation.

What you can do with it

  • Evaluating depth prediction models on sparse panoramic frames
  • Testing 3D reconstruction and novel view synthesis algorithms with minimal viewpoints
Tasks
Depth estimation, Image → image
Languages
English
Size
10K–100K rows
Downloads this month
21.2K
Likes
14

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("CM-EVS-Dataset/cmevs-erp-eval")