Skip to content
Back to dataset directory

Human demonstrations · Apple

EgoDex

Egocentric tabletop manipulation video with head, upper-body and hand poses.

Get datasetOfficial source

What is included?

Data type
Human demonstrations
Tasks
Human-to-robot transfer · Hand-object interaction
Modalities
RGB · Pose · Language
Embodiment
Human participants wearing Apple Vision Pro

Reported scale

829 video hours · 194 tasks

Publisher-reported scope. Preserve the stated unit; no duration or trajectory conversion is inferred.

Access & license

Follow the original publisher’s dataset download and loading instructions.

Dataset license not confirmed in this listing

Before you use this dataset

The repository license describes software; this review does not infer a dataset-wide license from it. Human pose annotations are not robot commands.

Publisher sources checked 2026-10-05. Listing review checks the documentation, not downloaded data quality. Always check the exact release and terms.