Table 1 Comparison of the proposed dataset with existing dataset on hand interaction.

Data	Modality	Egocentric	Hand	Multi-objects	Label
Data	Modality	Egocentric	Hand	Multi-objects	hand-object	action-segment
YALE⁸	RGB	×	×	×	×	×
ActionSense³⁴	RGB, Depth, Tactile, body	√	√	×	×	√
GRAB¹³	RGB, Depth, Body	√	√	×	√	×
ATTACH¹⁵	RGB, Body	×	√	×	×	√
ContactPose¹⁰	RGB, Depth	×	√	×	√	×
ARCTIC¹⁶	RGB, Depth	√	√	×	√	×
Hasson et al.¹²	Body	×	√	×	√	×
HOI4D³²	RGB, Depth	√	√	×	√	×
Hernando et al.²³	RGB, Depth	√	×	×	√	×
ObMan¹²	RGB	×	×	×	√	×
Dreher et al.¹¹	RGB	×	√	×	√	√
Sun et al.⁷	RGB	×	×	√	×	×
Elangovan et al.²⁸	RGB	√	√	×	×	×
KIT³¹	RGB, Tactile, Body	√	√	×	×	×
DexYCB²⁵	RGB, Depth	×	√	×	√	×
HOnnotate²⁴	RGB, Depth	×	×	×	√	×
HANDdata²⁹	Radar, Tof, IMU	×	×	×	×	×
ALOHA³⁰	RGB, Depth	×	√	√	×	×
Kaiwu Kitchen	RGB, Depth, Tactile, Body, IMU, Eye-tracking, EMG	√	√	√	√	√

It involves different modalities, individual viewpoints, presence of both hands, presence of multiple objects grasps, and labeling types.

Quick links

Search