Dataset · nosmall
Eval-Our panoramic depth benchmark
nosmall/eval_our
An evaluation dataset for panoramic depth from eight open sources.
What's inside
The dataset contains images and texts with a total volume of 15,766 test samples (ranging in size from 10K to 100K). It is a unified benchmark for evaluating the depth of equidistant (360°) panoramas, compiled from the test splits of eight panoramic datasets.
What you can do with it
- Evaluating panoramic image depth estimation models
- Benchmarking 3D computer vision
- Size
- 10K–100K rows
- Downloads this month
- 23.7K
- Likes
- 0
How to load it
from datasets import load_dataset
ds = load_dataset("nosmall/eval_our")