Video · Image · Multimodal / Education
Real-World Egocentric POV Human Task Video Dataset for Physical AI and Robotics
Advertised by providerLicensing rights not independently verifiedPaid licence
Overview
Real-World Egocentric POV Human Task Video Dataset for Physical AI and Robotics is a multimodal dataset. Typical applications include robotics & embodied AI. Relevant sectors: Education, AI & Machine Learning and Manufacturing & Industrial. The listing mentions timestamps. It is offered as a paid dataset; pricing details are provided on request.
Primary applications
What the listing offers
- Listing mentions timestamps.
This egocentric dataset is produced on demand based on the buyer’s required tasks, geography, participant profiles, capture setup, annotation depth and target volume. Pilot samples can be produced before full-scale collection. The dataset provides rights-cleared egocentric video of people performing complete, physically meaningful tasks in real-world environments. Recordings are captured from a head-mounted, chest-mounted or wearable first-person camera, with the participant’s hands, tools, manipulated objects and surrounding workspace clearly visible. The dataset is designed for Physical AI, embodied AI, humanoid robotics, vision-language-action models, robot imitation learning, world models, egocentric action recognition, human-object interaction analysis and long-horizon task understanding. Task coverage can include household activities, cooking, cleaning, furniture assembly, maintenance, retail operations, shelf stocking, picking and packing, sorting, scanning, labeling, warehouse workflows, construction support, plumbing, electrical work and other skilled professional activities. Data can be collected during authentic daily or professional work, allowing participants to record tasks they already perform, or through controlled task protocols when greater consistency, repetition and balanced coverage are required. Each delivery can include continuous first-person video, synchronized audio, session metadata, task descriptions, timestamps, camera and environment information,
Technical specifications
Related datasets
More in Video & Physical AI →Video · Image · MultimodalAdvertised
Embodied AI Data Solutions | Data Collection & Annotation | AI Robot | Multimodal Data | Video Data
1. Overview the provider provide accurate and comprehensive data support for various embodied intelligent applications.
- Language
- —
- Volume
- —
- Domain
- AI & Machine Learning +3
- Use case
- Alignment & RLHF
USD 10,000