gSDF: Geometry-Driven Signed Distance Functions for 3D Hand-Object Reconstruction

TL;DR

This paper introduces gSDF, a geometry-guided SDF approach for 3D hand-object reconstruction from monocular images, leveraging hand structure, pose estimation, and temporal data to improve accuracy and robustness.

Contribution

The work integrates hand kinematics with SDFs for detailed 3D reconstruction, addressing limitations of traditional SDFs by explicitly modeling geometry.

Findings

Significant improvements over state-of-the-art on ObMan and DexYCB benchmarks.

Effective handling of occlusion and motion blur through temporal information.

Enhanced 3D reconstruction quality with geometry alignment.

Abstract

Signed distance functions (SDFs) is an attractive framework that has recently shown promising results for 3D shape reconstruction from images. SDFs seamlessly generalize to different shape resolutions and topologies but lack explicit modelling of the underlying 3D geometry. In this work, we exploit the hand structure and use it as guidance for SDF-based shape reconstruction. In particular, we address reconstruction of hands and manipulated objects from monocular RGB images. To this end, we estimate poses of hands and objects and use them to guide 3D reconstruction. More specifically, we predict kinematic chains of pose transformations and align SDFs with highly-articulated hand poses. We improve the visual features of 3D points with geometry alignment and further leverage temporal information to enhance the robustness to occlusion and motion blurs. We conduct extensive experiments on the challenging ObMan and DexYCB benchmarks and demonstrate significant improvements of the proposed method over the state of the art.