arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2207.03386cs.RO
收藏

自主机器人动力学预测与适应的自我中心视觉自建模

Egocentric Visual Self-Modeling for Autonomous Robot Dynamics Prediction and Adaptation

Columbia University(哥伦比亚大学);Duke University(杜克大学)

Yuhang Hu, Boyuan Chen, Hod Lipson

AI总结:

本文提出一种仅用单目第一视角相机、自监督学习的任务无关动力学自模型,在12自由度机器人上验证了运动能力、异常检测与行为适应,并展示了跨配置的泛化性,为自主、适应和韧性机器人系统奠定基础。

摘要:

机器人对其自身动力学进行建模的能力,是自主规划与学习以及自主损伤检测与恢复的关键。传统上,动力学模型是预先编程或从外部观测中学习得到的。在此,我们首次展示了一种任务无关的动力学自模型,如何仅使用单个第一人称视角摄像头,以自监督方式学习,且无需任何关于机器人形态、运动学或任务的先验知识。通过在12自由度机器人上的实验,我们展示了该模型在基于视觉输入的基本运动任务中的能力。值得注意的是,机器人能够自主检测异常(如损坏部件),并调整其行为,展现了在动态环境中的韧性。此外,该模型的泛化能力在不同配置的机器人上得到了验证,强调了其作为通用工具在多样化机器人系统中的潜力。我们工作中提出的自我中心视觉自模型,为更自主、更适应性强、更具韧性的机器人系统铺平了道路。

英文摘要:

The ability of robots to model their own dynamics is key to autonomous planning and learning, as well as for autonomous damage detection and recovery. Traditionally, dynamic models are pre-programmed or learned from external observations. Here, we demonstrate for the first time how a task-agnostic dynamic self-model can be learned using only a single first-person-view camera in a self-supervised manner, without any prior knowledge of robot morphology, kinematics, or task. Through experiments on a 12-DoF robot, we demonstrate the capabilities of the model in basic locomotion tasks using visual input. Notably, the robot can autonomously detect anomalies, such as damaged components, and adapt its behavior, showcasing resilience in dynamic environments. Furthermore, the model's generalizability was validated across robots with different configurations, emphasizing its potential as a universal tool for diverse robotic systems. The egocentric visual self-model proposed in our work paves the way for more autonomous, adaptable, and resilient robotic systems.

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。

↑