Ego-Centric Dataset
First-person perspective video data captured via wearable cameras. Records how humans perceive and interact with their surroundings — providing essential training data for embodied AI, activity understanding, and robot learning systems.
Key Features
Data Collection Setup
Wearable Camera-Based Collection
The dataset is collected using wearable devices — GoPro cameras and smartphones — to capture activities from the user’s natural viewpoint in real-world environments.
Dataset Coverage
Real-World Human Activities
Diverse human activities across multiple environments:
Example
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Manicure / Nail Art
A human performs a full manicure and nail art routine, captured from a first-person wearable camera as a continuous 16-frame sequence showing the complete process step by step.
Captures: Fine hand-tool interaction · Nail shaping & polishing · Step-by-step process detail · Precision manipulation
Process sequence
16 frames · ego view12345678910111213141516Dataset Specifications
Applications
Designed for Embodied AI Development
Human Activity Recognition Action Understanding Imitation Learning VLA Training Robot Learning AI Agent Training
Interested in using this dataset for your research or product development? Get in touch with our team.