Buy licensed first-person (egocentric) video
Robotics teams and AI labs training agents on manual and screen-based work need footage from the worker's point of view, with the rights in order. Tell us the tasks, the hours and the setting. We find sellers who hold matching footage, check the rights, and send a summary and a cleaned sample that the seller has approved.
New to the format? Read what first-person video is. Buying records instead? See how buying company data works.
What to tell us
| Choose | What to specify |
|---|---|
| Tasks and tools | Which hands-on or screen-based tasks, and which tools or software |
| Settings | Kitchens, shops, job sites, warehouses, offices, and the countries you can accept |
| Volume | Hours of footage, and how many different people you need filmed |
| Capture | Camera type and position, resolution, frame rate, and whether you need audio |
| Labels | Narration, action segments, hand trajectories, keypoints, object grounding, outcomes or reviewer notes, if you need them |
| Regions and languages | The country the footage was recorded in and the language of any narration |
| Rights | Use limits, how long, and whether you want an exclusive agreement for a term |
You do not need all of it to start. A task and a rough volume is enough.
What you receive
- A one-page summary of the set: hours, people, tasks, settings and capture details
- A cleaned sample the seller has approved
- A written note on ownership and consent
- The full set after signing, with a provenance memo and an acceptance window
What we check first
- The seller owns the footage outright
- Each wearer agreed in writing to this use
- People in the background are blurred, muted or cut
- The footage was not taken in secret or from the web
How buying works
Tell us what you need
Data type, industry, volume, exclusive AI lab agreement or not. Free to ask.
We source and clear rights
We match sellers and check ownership first.
Review, then quote
A one-page summary and de-identified samples, then your quote.
Sign, then receive
Terms in writing. The package comes with a provenance memo.
Accept, then pay
Within the written acceptance window.
The full steps are on how buying works.
More on first-person video
- Sell first-person video. What you need to hold, and how a deal runs.
- Video for robotics and physical AI. What models learn from it and the tasks labs ask for.
- Egocentric video annotation. The labels buyers ask for, and who adds them.
Questions about buying first-person video
What can you source?
Egocentric video of hands-on and screen-based work from sellers who hold the footage and the rights. What exists depends on the task and the setting you need, so tell us and we come back with what is available and what we can find.
Do you hold the footage?
No. The seller keeps it. You receive a summary and a cleaned sample the seller has approved. The full set is delivered after signing, through us.
How do you handle consent?
Before any summary goes out we confirm who owns the footage, that each wearer agreed in writing to this use, and how other people in frame were handled. You get that in writing with the package.
Can we ask for a specific task or setting?
Yes. Say which tasks, tools and environments, how many hours and how many different people you need. The more specific the request, the better the match.
Can we get annotations or narration?
Ask for it. Some sellers keep step lists, narration or reviewer notes with their clips, and we flag them. Annotation done after the fact is agreed case by case.
Can we get an exclusive agreement?
An exclusive agreement is possible for a fixed term, agreed case by case and priced. Use rights are usually perpetual. One with no end date is not offered.
Didn't find your question? Ask us directly.