Dataset · benediktkol
Synthetic drone images with depth maps
benediktkol/DDOS
Synthetic aerial drone images with depth maps and per-pixel obstacle masks, including wires.
What's inside
The dataset contains 10 to 100 thousand synthetic aerial images, supplemented with depth maps and semantic segmentation masks. Special emphasis in labeling is placed on detecting thin objects such as wires. Specific file format information is not provided in the card.
What you can do with it
- Training obstacle and thin-object segmentation models for UAVs
- Estimating scene depth from onboard drone camera images
- Tasks
- Image segmentation, Depth estimation
- Size
- 10K–100K rows
- Downloads this month
- 193.9K
- Likes
- 11
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("benediktkol/DDOS")