Dataset · ropedia-ai
3D scenes and first-person videos for robotics
ropedia-ai/Ropedia-Xperience-2
A large Physical AI dataset with 3D room scans and egocentric recordings in a unified metric system.
What's inside
A large-scale dataset for robotics, 3D reconstruction, and multimodal analysis, fully including over 1,500 scanned scenes and over 4,000 first-person recordings in a unified coordinate system. The first part of 126 scenes, sized at about 1.07 TB, has been published, including images, tabular data, and English text descriptions.
What you can do with it
- Train spatial perception and navigation models for robots and embodied agents
- Train algorithms for image-to-3D scene reconstruction and keypoint detection
- Train multimodal video-text-to-text models to understand actions in physical space
- Tasks
- Robotics, Image → 3d, Keypoint detection, Video text → text
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 8.8K
- Likes
- 63
How to load it
from datasets import load_dataset
ds = load_dataset("ropedia-ai/Ropedia-Xperience-2")