arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.01153cs.RO
收藏

UniForce: 一种用于具有多样化触觉传感器的机器人操作的统一潜在力模型

UniForce: A Unified Latent Force Model for Robot Manipulation with Diverse Tactile Sensors

Zhuo Chen, Fei Ni, Kaiyao Luo, Zhiyuan Wu, Xuyang Zhang, Emmanouil Spyrakos-Papastavridis, Lorenzo Jamone, Nathan F. Lepora, Jiankang Deng, Shan Luo

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

UniForce提出一种统一的触觉表示学习框架，通过联合建模逆正动力学和力平衡损失，实现跨不同触觉传感器的力感知任务通用化，提升机器人操作的力感知能力。

摘要：

力感知对于灵巧的机器人操作至关重要，但扩展力感知策略学习受到触觉传感器异质性的影响。差异在传感原理（例如光学与磁性）、外形尺寸和材料通常需要特定传感器的数据收集、校准和模型训练，从而限制了通用性。我们提出了UniForce，一种新的统一触觉表示学习框架，该框架在多样化的触觉传感器上学习共享的潜在力空间。UniForce通过联合建模逆动力学（图像到力）和正动力学（力到图像），并受力平衡和图像重建损失的约束，以生成力基础的表示。为了避免依赖昂贵的外部力/扭矩（F/T）传感器，我们利用静态平衡并收集力配对数据，通过直接的传感器-物体-传感器相互作用，使跨传感器对齐与接触力。由此产生的通用触觉编码器可以无缝插入到下游的力感知机器人操作任务中，无需重新训练或微调。在异构触觉传感器（包括GelSight、TacTip和uSkin）上的广泛实验表明，与先前方法相比，力估计有持续的改进，并在视觉-触觉-语言-动作（VTLA）模型中实现了有效的跨传感器协调，用于机器人擦除任务。代码和数据集将被发布。

英文摘要：

Force sensing is essential for dexterous robot manipulation, but scaling force-aware policy learning is hindered by the heterogeneity of tactile sensors. Differences in sensing principles (e.g., optical vs. magnetic), form factors, and materials typically require sensor-specific data collection, calibration, and model training, thereby limiting generalisability. We propose UniForce, a novel unified tactile representation learning framework that learns a shared latent force space across diverse tactile sensors. UniForce reduces cross-sensor domain shift by jointly modeling inverse dynamics (image-to-force) and forward dynamics (force-to-image), constrained by force equilibrium and image reconstruction losses to produce force-grounded representations. To avoid reliance on expensive external force/torque (F/T) sensors, we exploit static equilibrium and collect force-paired data via direct sensor--object--sensor interactions, enabling cross-sensor alignment with contact force. The resulting universal tactile encoder can be plugged into downstream force-aware robot manipulation tasks with zero-shot transfer, without retraining or finetuning. Extensive experiments on heterogeneous tactile sensors including GelSight, TacTip, and uSkin, demonstrate consistent improvements in force estimation over prior methods, and enable effective cross-sensor coordination in Vision-Tactile-Language-Action (VTLA) models for a robotic wiping task. Code and datasets will be released.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑