Stereo Hand-Object Reconstruction for Human-to-Robot Handover

TL;DR

This paper introduces a stereo-based hand-object reconstruction method that combines learned 3D priors and RGB inputs to improve human-to-robot handover, especially for transparent and unseen objects.

Contribution

It presents a novel stereo reconstruction approach using learned shape priors and RGB data, enhancing generalization and handling transparent objects in human-to-robot handovers.

Findings

Reduces object Chamfer distance compared to existing methods.

Enables successful robot handover of diverse household objects.

Improves reconstruction accuracy in stereo and single-view settings.

Abstract

Jointly estimating hand and object shape facilitates the grasping task in human-to-robot handovers. However, relying on hand-crafted prior knowledge about the geometric structure of the object fails when generalising to unseen objects, and depth sensors fail to detect transparent objects such as drinking glasses. In this work, we propose a stereo-based method for hand-object reconstruction that combines single-view reconstructions probabilistically to form a coherent stereo reconstruction. We learn 3D shape priors from a large synthetic hand-object dataset to ensure that our method is generalisable, and use RGB inputs to better capture transparent objects. We show that our method reduces the object Chamfer distance compared to existing RGB based hand-object reconstruction methods on single view and stereo settings. We process the reconstructed hand-object shape with a projection-based outlier removal step and use the output to guide a human-to-robot handover pipeline with wide-baseline stereo RGB cameras. Our hand-object reconstruction enables a robot to successfully receive a diverse range of household objects from the human.