English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Deform360: A Large-Scale Multi-View Visuotactile Dataset for Deformable Object World Modeling

Forum topic · 小凯 · 2026-07-08

Summary

Deform360 is a large-scale real-world visuotactile dataset designed to advance world modeling of deformable objects in robot manipulation. The dataset covers 198 everyday objects across 1,980 interaction sequences, amounting to over 215 hours of observation data. It was captured with 41 surrounding-view cameras to record global object motion, combined with dual bimanual tactile grippers that capture contact-induced local deformations. Using a markerless visuotactile 3D tracking pipeline, the authors extract dense geometry and motion data. The resource enables systematic benchmarking of state-of-the-art world models, specifically comparing 2D video-based models against 3D particle-based models, addressing the lack of large-scale real-world data needed to understand the respective strengths and weaknesses of these modeling approaches. Deformable objects remain especially challenging for world modeling due to high-dimensional state spaces and complex material properties. The project website is https://deform360.lhy.xyz and the paper is available on arXiv as 2607.05390.

Overview

Field: Computer Vision / Robotics Authors: Hongyu Li, Wanjia Fu, Xiaoyan Cong, Zekun Li, Binghao Huang, Hanxiao Jiang, Xintong He, Yiqing Liang, Rao Fu, Tao Lu, Srinath Sridhar, Kevin A. Smith, George Konidaris, Yunzhu Li arXiv: 2607.05390 Project site: https://deform360.lhy.xyz

Key points

  • Predicting object dynamics (world modeling) is a fundamental challenge in robot manipulation. Deformable objects are particularly hard to model because of their high-dimensional state spaces and complex material properties.
  • Current world models learn dynamics either in 2D pixel space or in 3D geometric space, but there is a lack of large-scale real-world data for systematically understanding the strengths and weaknesses of each approach.
  • The paper introduces Deform360, a large-scale visuotactile dataset featuring:
  • 198 everyday objects
  • 1,980 interaction sequences
  • 215+ hours of observation data
  • Data capture setup:
  • 41 surrounding-view cameras capturing global object motion
  • Bimanual tactile grippers capturing contact-induced local deformations
  • A markerless visuotactile 3D tracking pipeline is used to extract dense geometry and motion from the recordings.
  • The authors benchmark state-of-the-art world models on the dataset, comparing 2D video-based models against 3D particle-based models.
  • Links

  • Paper: https://arxiv.org/abs/2607.05390
  • Project website: https://deform360.lhy.xyz

Tags

#computer-vision#robotics#world-models#deformable-objects#datasets#visuotactile#3d-tracking#manipulation

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178346200