InterHand-2.6M

The 'InterHand2.6M' dataset is a benchmark that contains a large collection of images for evaluating the 3D pose estimation of two hands in close interaction, addressing challenges such as occlusions and joint confusions.

Papers using InterHand-2.6M (2)

Interacting Attention Graph For Single Image Two-hand Reconstruction2022 · 110 citesKeypoint Transformer: Solving Joint Identification In Challenging Hands And Object Interactions For Accurate 3D Pose Estimation2021 · 9 cites