GuildSign in
Back to marketplace

Person gesturing with hands in indoor setting

2/5, 1 min

Evidence trail

Model and reviewer agree on 0 of 3 labels

What was measured, what the model proposed, what a person changed, and why buyers took it. Nothing here can be edited after the fact.

  1. Observed on the device

    Resolution: 1280x720, Length: 15s, Lighting: Good, Sharpness: Sharp, Camera steadiness: Steady, Hands in frame: 0%

    Seller entered: no labels

    Filmed live in the app: challenge passed (show 3 fingers), hand tracking covers 100%, motion sensors not available on this device

  2. Open-source models labelled it on the device

    Objects: umbrella 78%

  3. Model proposed (kimi-k2.6)

    Labels: Software task, Office, Egocentric

    Skill: unclear. No tool handling or technical task is visible; only casual hand gestures toward the camera.

    Content score 1/5. No manipulation task or physical work is visible No tools, materials, or software interface shown Person appears to be talking or gesturing to camera rather than performing a demonstrable task

    Flags: faces

  4. Scored and priced

    Score 2/5 from technical 62%, labels 0%, model 1/5

    Market rate $5/h, listed at $0.02

Who filmed it

Mr. Wong. Electrician, 14 years. Master

Credential (self-declared, not yet checked): Registered electrical worker

How it was captured, and the licence

Recorded inside the Guild app, with a live hand-motion trace.

  • The contributor keeps ownership of the footage.
  • Non-exclusive: the same clip can be licensed to other buyers.
  • The buyer may train and evaluate commercial robotics and embodied AI models on it.
  • The buyer may not resell, sublicense or publish the footage itself.
  • Not licensed for surveillance, biometric identification, or generating likenesses of the people shown.
  • Perpetual for models already trained. If the contributor withdraws the clip, no new licences are sold.

Steps

  1. person raises both hands with palms facing camera
  2. person moves hands slightly while facing camera
  3. person lowers left hand and gestures with right hand
  4. person continues gesturing with right hand