arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.01488cs.ROcs.AIcs.LG
收藏

自我改进机器人:端到端自主视觉强化学习

Self-Improving Robots: End-to-End Autonomous Visuomotor Reinforcement Learning

Stanford University(斯坦福大学)

Archit Sharma, Ahmed M. Ahmed, Rehaan Ahmad, Chelsea Finn

摘要:

在模仿和强化学习中,人类监督的成本限制了机器人可以训练的数据量。一个令人向往的目标是构建自我改进的机器人:可以在最少的人类监督或监督下自主互动的机器人。这样的机器人可以收集和训练更大的数据集,从而学习更强大和性能更好的政策。虽然强化学习通过试错法为这种自主学习提供了一个框架,但实际的实现最终需要广泛的人类监督来设计奖励函数,并在交互事件之间重复重置环境。在这项工作中,我们提出了MEDAL++,一种新的设计,用于自我改进的机器人系统:在开始时给出一小组专家演示,机器人通过学习做和撤销任务来自主练习任务,同时从演示中推断奖励函数。策略和奖励函数是从高维视觉输入端到端学习的,绕过了对先前工作中使用的视觉编码器的显式状态估计或特定于任务的预训练的需要。我们首先在模拟的非情景基准EARL上评估了我们提出的算法,发现MEDAL++的数据效率更高,与最先进的基于视觉的方法相比,最终性能提高了30%。我们的真实机器人实验表明,MEDAL++可以应用于操作问题,在更大的环境比那些考虑在以前的工作,和自主自我改进可以提高成功率30-70%以上的行为克隆只是专家数据。代码、培训和评估视频以及简要概述请访问:https://architsharma97.github.io/self-improving-robots/

英文摘要:

In imitation and reinforcement learning, the cost of human supervision limits the amount of data that robots can be trained on. An aspirational goal is to construct self-improving robots: robots that can learn and improve on their own, from autonomous interaction with minimal human supervision or oversight. Such robots could collect and train on much larger datasets, and thus learn more robust and performant policies. While reinforcement learning offers a framework for such autonomous learning via trial-and-error, practical realizations end up requiring extensive human supervision for reward function design and repeated resetting of the environment between episodes of interactions. In this work, we propose MEDAL++, a novel design for self-improving robotic systems: given a small set of expert demonstrations at the start, the robot autonomously practices the task by learning to both do and undo the task, simultaneously inferring the reward function from the demonstrations. The policy and reward function are learned end-to-end from high-dimensional visual inputs, bypassing the need for explicit state estimation or task-specific pre-training for visual encoders used in prior work. We first evaluate our proposed algorithm on a simulated non-episodic benchmark EARL, finding that MEDAL++ is both more data efficient and gets up to 30% better final performance compared to state-of-the-art vision-based methods. Our real-robot experiments show that MEDAL++ can be applied to manipulation problems in larger environments than those considered in prior work, and autonomous self-improvement can improve the success rate by 30-70% over behavior cloning on just the expert data. Code, training and evaluation videos along with a brief overview is available at: https://architsharma97.github.io/self-improving-robots/

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。

↑