arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.04852cs.RO
收藏

基于自监督环境综合的真实世界自主导航学习

Learning Real-world Autonomous Navigation by Self-Supervised Environment Synthesis

The University of Texas at Austin(德克萨斯大学奥斯汀分校);George Mason University(乔治梅森大学);Sony AI(索尼人工智能)

Zifan Xu, Anirudh Nair, Xuesu Xiao, Peter Stone

摘要:

机器学习方法最近已经以数据驱动的方式实现了移动机器人的自主导航。由于大多数现有的基于学习的导航系统都是使用人工创建的训练环境中生成的数据进行训练的,因此在大规模的真实世界部署过程中,机器人将不可避免地遇到不可见的场景,这些场景不在训练分布中,因此导致真实世界性能不佳。另一方面,在现实世界中直接训练通常是不安全和低效的。为了解决这个问题,我们引入了自监督环境合成(SES),其中,在具有安全性和效率要求的真实世界部署之后,自主移动机器人可以利用来自真实世界部署的经验,重建导航场景,并在仿真中合成具有代表性的训练环境。在这些合成环境中进行训练可以提高未来在现实世界中的表现。SES在合成代表性的仿真环境和提高现实世界的导航性能的有效性进行评估,通过大规模部署在一个高保真,逼真的模拟器和小规模部署的物理机器人。

英文摘要:

Machine learning approaches have recently enabled autonomous navigation for mobile robots in a data-driven manner. Since most existing learning-based navigation systems are trained with data generated in artificially created training environments, during real-world deployment at scale, it is inevitable that robots will encounter unseen scenarios, which are out of the training distribution and therefore lead to poor real-world performance. On the other hand, directly training in the real world is generally unsafe and inefficient. To address this issue, we introduce Self-supervised Environment Synthesis (SES), in which, after real-world deployment with safety and efficiency requirements, autonomous mobile robots can utilize experience from the real-world deployment, reconstruct navigation scenarios, and synthesize representative training environments in simulation. Training in these synthesized environments leads to improved future performance in the real world. The effectiveness of SES at synthesizing representative simulation environments and improving real-world navigation performance is evaluated via a large-scale deployment in a high-fidelity, realistic simulator and a small-scale deployment on a physical robot.

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。

↑