arXiv cs.AI· Ri-Zhao Qiu, Bowen Wen, Yifeng Jiang, Chen Tessler, Eric Heiden, Xue Bin Peng, Xiaolong Wang, Stan Birchfield, Erwin Coumans·· 13 小时前AI 评分34
ActiveAlign：用主动感知对齐第一人称视角操作中的 Real2Sim2Real 差距
ActiveAlign: Aligning Egocentric Manipulation via Active Perception
AI 导读
ActiveAlign 框架通过主动感知对齐 Real2Sim2Real 差距，包含 ActiveAlign-Waving（利用主动手部运动与可微分渲染完成手眼标定）和 ActiveAlign-Touching（借助触觉反馈将物体对齐精度提升至毫米级）两个模块。该框架可用于为开源未标定的机器人操作数据集整理仿真环境，并在真实 Unitree G1 机器人上实现了按压按钮、移动滑块等毫米级精细操作。
正文 · 原文
Abstract:Real-to-Sim-to-Real (Real2Sim2Real) enables structured world representations for planning, interpretability, and safe execution. Most existing approaches rely on passive perception, which creates real2sim gaps due to imperfect observation. Inspired by how humans perceive actively through actions, we propose ActiveAlign, a framework that leverages active perception to align the Real2Sim2Real gap from interactions. ActiveAlign introduces two complementary modules: (i) ActiveAlign-Waving, which performs hand-eye calibration by exploiting active hand motions and differentiable rendering, and (ii) ActiveAlign-Touching, which refines object alignment through tactile feedback to achieve millimeter-level precision. Together, these capabilities enable strong real2sim alignment. We demonstrate that ActiveAlign-Waving can be used to curate simulation environments for open-sourced uncalibrated robot manipulation datasets, and demonstrate on a real Unitree G1 robot that ActiveAlign enables fine manipulation of millimeter-scale objects such as pressing buttons and moving sliders.
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2610.36859 [cs.RO] |
| (or arXiv:2610.36859v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2610.36859 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ri-Zhao Qiu [view email]
[v1]
Tue, 27 Oct 2026 08:54:27 UTC (7,337 KB)
来源：arXiv cs.AI · arxiv.org