WHOLE: World-Grounded Hand-Object Lifted from Egocentric Videos

TL;DR

WHOLE is a novel method that jointly reconstructs hand and object motion from egocentric videos, significantly improving accuracy in challenging occlusion and out-of-view scenarios by learning a generative prior for their interactions.

Contribution

It introduces a holistic approach with a learned generative prior to jointly reason about hand-object interactions in world space from egocentric videos.

Findings

State-of-the-art performance in hand motion estimation

Superior 6D object pose estimation accuracy

Effective reconstruction of hand-object interactions

Abstract

Egocentric manipulation videos are highly challenging due to severe occlusions during interactions and frequent object entries and exits from the camera view as the person moves. Current methods typically focus on recovering either hand or object pose in isolation, but both struggle during interactions and fail to handle out-of-sight cases. Moreover, their independent predictions often lead to inconsistent hand-object relations. We introduce WHOLE, a method that holistically reconstructs hand and object motion in world space from egocentric videos given object templates. Our key insight is to learn a generative prior over hand-object motion to jointly reason about their interactions. At test time, the pretrained prior is guided to generate trajectories that conform to the video observations. This joint generative reconstruction substantially outperforms approaches that process hands and objects separately followed by post-processing. WHOLE achieves state-of-the-art performance on hand motion estimation, 6D object pose estimation, and their relative interaction reconstruction. Project website: https://judyye.github.io/whole-www