Skip to content

Wake collections

Kitchen and trade video dataset.

Inspect first-person kitchen and trade activity, with attention to tool handling, preparation order and changes in the work surface.
Kitchen and trade video dataset. Collection recording still.
Kitchen workflows and skilled trades

Open the kitchen and trade video dataset. collection

Follow the preparation sequence.

Look for the order of actions, tool changes and visible ingredient or material states. Occlusion near the hands can hide the step a model needs to learn.

Label observable transitions.

Define when preparation begins, when an item changes state and what counts as a completed step. Separate interruptions from the action itself.

Scope a workplace recording.

Name the station, tools, ingredients or materials, and workflow variations. Agree access and permissions before recording workers or customers.

Questions, answered.

What labels matter for kitchen activity?

Task instructions, tool and object names, action boundaries and visible state changes help describe a preparation sequence. Confirm which labels exist in the selected collection.

Can recordings be used directly as robot commands?

No. First-person kitchen video does not itself provide a robot control signal.

Keep exploring.

What should your robot learn next?

Bring the task, setting and signals you need. We’ll work through the collection and annotation requirements with you.

Build with Wake