Research

Paper

TESTING March 11, 2026

TacLoc: Global Tactile Localization on Objects from a Registration Perspective

Authors

Zirui Zhang, Boyang Zhang, Fumin Zhang, Huan Yin

Abstract

Pose estimation is essential for robotic manipulation, particularly when visual perception is occluded during gripper-object interactions. Existing tactile-based methods generally rely on tactile simulation or pre-trained models, which limits their generalizability and efficiency. In this study, we propose TacLoc, a novel tactile localization framework that formulates the problem as a one-shot point cloud registration task. TacLoc introduces a graph-theoretic partial-to-full registration method, leveraging dense point clouds and surface normals from tactile sensing for efficient and accurate pose estimation. Without requiring rendered data or pre-trained models, TacLoc achieves improved performance through normal-guided graph pruning and a hypothesis-and-verification pipeline. TacLoc is evaluated extensively on the YCB dataset. We further demonstrate TacLoc on real-world objects across two different visual-tactile sensors.

Metadata

arXiv ID: 2603.10565
Provider: ARXIV
Primary Category: cs.RO
Published: 2026-03-11
Fetched: 2026-03-12 04:21

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.10565v1</id>\n    <title>TacLoc: Global Tactile Localization on Objects from a Registration Perspective</title>\n    <updated>2026-03-11T09:15:07Z</updated>\n    <link href='https://arxiv.org/abs/2603.10565v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.10565v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Pose estimation is essential for robotic manipulation, particularly when visual perception is occluded during gripper-object interactions. Existing tactile-based methods generally rely on tactile simulation or pre-trained models, which limits their generalizability and efficiency. In this study, we propose TacLoc, a novel tactile localization framework that formulates the problem as a one-shot point cloud registration task. TacLoc introduces a graph-theoretic partial-to-full registration method, leveraging dense point clouds and surface normals from tactile sensing for efficient and accurate pose estimation. Without requiring rendered data or pre-trained models, TacLoc achieves improved performance through normal-guided graph pruning and a hypothesis-and-verification pipeline. TacLoc is evaluated extensively on the YCB dataset. We further demonstrate TacLoc on real-world objects across two different visual-tactile sensors.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.RO'/>\n    <published>2026-03-11T09:15:07Z</published>\n    <arxiv:comment>8 pages, 12 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.RO'/>\n    <author>\n      <name>Zirui Zhang</name>\n    </author>\n    <author>\n      <name>Boyang Zhang</name>\n    </author>\n    <author>\n      <name>Fumin Zhang</name>\n    </author>\n    <author>\n      <name>Huan Yin</name>\n    </author>\n  </entry>"
}