Datasets
VideoRobotics
Hands-Only Manipulation
Filmed at close range with the focus tight on the hands, hands-only manipulation footage isolates the dexterous interaction between fingers and objects. The form covers everyday object handling, tool use, fine assembly, packaging, food preparation, and craft work, all framed to keep grip, contact points, and finger articulation in clear view.
Hours11.8K+
Objects100+
Activities8+
Training Use Cases
✓Dexterous manipulation policy learning
✓Robot imitation learning from human demonstrations
✓Hand-object interaction modeling for AR and VR
✓Video-conditioned manipulation generation
Key Highlights
✓100+ object categories handled including hand tools, kitchen utensils, electronics, packaging, fasteners, and craft materials
✓8+ activity contexts covered including food prep, assembly, packaging, fastening, tool use, and craft work
✓Tight close-up framing throughout, keeping finger and grip detail in clear view
✓Both single-hand and two-handed (bimanual) sequences covered
Metadata Fields
durationLength of clip in HH:MM:SS
resolutionPixel dimensions (e.g., 1920x1080, 3840x2160)
frame_rateFrames per second (e.g., 24, 30, 60, 120)
contains_audioWhether the clip carries an audio track (boolean)
primary_categoryDominant content category assigned to a video
object_classhand_tool | kitchen_utensil | electronics | packaging | fastener | craft_material
activity_contextfood_prep | assembly | packaging | fastening | tool_use | craft_work
hand_count1 | 2