Research

Paper

AI LLM March 19, 2026

GHOST: Fast Category-agnostic Hand-Object Interaction Reconstruction from RGB Videos using Gaussian Splatting

Authors

Ahmed Tawfik Aboukhadra, Marcel Rogge, Nadia Robertini, Abdalla Arafa, Jameel Malik, Ahmed Elhayek, Didier Stricker

Abstract

Understanding realistic hand-object interactions from monocular RGB videos is essential for AR/VR, robotics, and embodied AI. Existing methods rely on category-specific templates or heavy computation, yet still produce physically inconsistent hand-object alignment in 3D. We introduce GHOST (Gaussian Hand-Object Splatting), a fast, category-agnostic framework for reconstructing dynamic hand-object interactions using 2D Gaussian Splatting. GHOST represents both hands and objects as dense, view-consistent Gaussian discs and introduces three key innovations: (1) a geometric-prior retrieval and consistency loss that completes occluded object regions, (2) a grasp-aware alignment that refines hand translations and object scale to ensure realistic contact, and (3) a hand-aware background loss that prevents penalizing hand-occluded object regions. GHOST achieves complete, physically consistent, and animatable reconstructions from a single RGB video while running an order of magnitude faster than prior category-agnostic methods. Extensive experiments on ARCTIC, HO3D, and in-the-wild datasets demonstrate state-of-the-art accuracy in 3D reconstruction and 2D rendering quality, establishing GHOST as an efficient and robust solution for realistic hand-object interaction modeling. Code is available at https://github.com/ATAboukhadra/GHOST.

Metadata

arXiv ID: 2603.18912
Provider: ARXIV
Primary Category: cs.CV
Published: 2026-03-19
Fetched: 2026-03-20 06:02

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.18912v1</id>\n    <title>GHOST: Fast Category-agnostic Hand-Object Interaction Reconstruction from RGB Videos using Gaussian Splatting</title>\n    <updated>2026-03-19T13:50:19Z</updated>\n    <link href='https://arxiv.org/abs/2603.18912v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.18912v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Understanding realistic hand-object interactions from monocular RGB videos is essential for AR/VR, robotics, and embodied AI. Existing methods rely on category-specific templates or heavy computation, yet still produce physically inconsistent hand-object alignment in 3D. We introduce GHOST (Gaussian Hand-Object Splatting), a fast, category-agnostic framework for reconstructing dynamic hand-object interactions using 2D Gaussian Splatting. GHOST represents both hands and objects as dense, view-consistent Gaussian discs and introduces three key innovations: (1) a geometric-prior retrieval and consistency loss that completes occluded object regions, (2) a grasp-aware alignment that refines hand translations and object scale to ensure realistic contact, and (3) a hand-aware background loss that prevents penalizing hand-occluded object regions. GHOST achieves complete, physically consistent, and animatable reconstructions from a single RGB video while running an order of magnitude faster than prior category-agnostic methods. Extensive experiments on ARCTIC, HO3D, and in-the-wild datasets demonstrate state-of-the-art accuracy in 3D reconstruction and 2D rendering quality, establishing GHOST as an efficient and robust solution for realistic hand-object interaction modeling. Code is available at https://github.com/ATAboukhadra/GHOST.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n    <published>2026-03-19T13:50:19Z</published>\n    <arxiv:primary_category term='cs.CV'/>\n    <author>\n      <name>Ahmed Tawfik Aboukhadra</name>\n    </author>\n    <author>\n      <name>Marcel Rogge</name>\n    </author>\n    <author>\n      <name>Nadia Robertini</name>\n    </author>\n    <author>\n      <name>Abdalla Arafa</name>\n    </author>\n    <author>\n      <name>Jameel Malik</name>\n    </author>\n    <author>\n      <name>Ahmed Elhayek</name>\n    </author>\n    <author>\n      <name>Didier Stricker</name>\n    </author>\n  </entry>"
}