arXiv 2409.194992025-02-01cs.RO

FastUMI：一种可扩展且硬件无关的通用操作接口及数据集

FastUMI: A Scalable and Hardware-Independent Universal Manipulation Interface with Dataset

Shanghai AI Lab（上海人工智能实验室）；Shanghai Jiao Tong University（上海交通大学）；University of Bristol（布里斯托大学）；Fudan University（复旦大学）；The University of Hong Kong（香港大学）；Xi’an Jiaotong-Liverpool University（西交利物浦大学）；Institute of AI, China Telecom Corp Ltd（中国电信股份有限公司人工智能研究院）

Zhaxizhuoma, Kehui Liu, Chuyue Guan, Zhongjie Jia, Ziniu Wu, Xin Liu, Tianyu Wang, Shuai Liang, Pengan Chen, Pingrui Zhang, Haoming Song, Delin Qu, Dong Wang, Z… 展开作者

Zhaxizhuoma, Kehui Liu, Chuyue Guan, Zhongjie Jia, Ziniu Wu, Xin Liu, Tianyu Wang, Shuai Liang, Pengan Chen, Pingrui Zhang, Haoming Song, Delin Qu, Dong Wang, Zhigang Wang, Nieqing Cao, Yan Ding, Bin Zhao, Xuelong Li

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出 FastUMI，一种可扩展且硬件无关的通用操作接口，通过解耦硬件设计与优化算法流程降低数据采集成本，并开源了超万条轨迹的数据集，推进了可扩展的数据驱动机器人学习。

摘要：

涉及机械臂的真实世界操作数据对于开发通用动作策略至关重要，然而此类数据仍然稀缺，因为现有的数据收集方法受到高成本、硬件依赖和复杂设置要求的阻碍。在这项工作中，我们介绍了 FastUMI，这是对 Universal Manipulation Interface (UMI) 系统的实质性重新设计，它通过实现快速部署、简化软硬件集成以及在真实世界数据采集中提供稳健性能来解决这些挑战。与 UMI 相比，FastUMI 具有几个优势：1) 它采用解耦的硬件设计并进行了广泛的机械修改，消除了对专用机器人组件的依赖，同时保持了一致的观察视角。2) 它还通过用现成的跟踪模块替换复杂的 Visual-Inertial Odometry (VIO) 实现来优化算法流程，在保持准确性的同时显著降低了部署复杂性。3) FastUMI 包含一个用于数据收集、验证以及与现有和新开发的模仿学习算法集成的生态系统，加速了策略学习的进步。此外，我们开源了一个包含超过 10,000 条真实世界演示轨迹的高质量数据集，涵盖 22 项日常任务，形成了迄今为止最多样化的类 UMI 数据集之一。实验结果证实，FastUMI 促进了快速部署，降低了运营成本和劳动力需求，并在多样化的操作场景中保持稳健的性能，从而推进了可扩展的数据驱动机器人学习。

英文摘要：

Real-world manipulation data involving robotic arms is crucial for developing generalist action policies, yet such data remains scarce since existing data collection methods are hindered by high costs, hardware dependencies, and complex setup requirements. In this work, we introduce FastUMI, a substantial redesign of the Universal Manipulation Interface (UMI) system that addresses these challenges by enabling rapid deployment, simplifying hardware-software integration, and delivering robust performance in real-world data acquisition. Compared with UMI, FastUMI has several advantages: 1) It adopts a decoupled hardware design and incorporates extensive mechanical modifications, removing dependencies on specialized robotic components while preserving consistent observation perspectives. 2) It also refines the algorithmic pipeline by replacing complex Visual-Inertial Odometry (VIO) implementations with an off-the-shelf tracking module, significantly reducing deployment complexity while maintaining accuracy. 3) FastUMI includes an ecosystem for data collection, verification, and integration with both established and newly developed imitation learning algorithms, accelerating policy learning advancement. Additionally, we have open-sourced a high-quality dataset of over 10,000 real-world demonstration trajectories spanning 22 everyday tasks, forming one of the most diverse UMI-like datasets to date. Experimental results confirm that FastUMI facilitates rapid deployment, reduces operational costs and labor demands, and maintains robust performance across diverse manipulation scenarios, thereby advancing scalable data-driven robotic learning.

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。