arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2202.07600cs.ROcs.LG
收藏

端到端移动操作的贝叶斯模仿学习

Bayesian Imitation Learning for End-to-End Mobile Manipulation

UC Berkeley(加州大学伯克利分校);Everyday Robots;Google Research(谷歌研究院)

Yuqing Du, Daniel Ho, Alexander A. Alemi, Eric Jang, Mohi Khansari

摘要:

在这项工作中,我们调查和证明贝叶斯方法的好处,从多个传感器输入的模仿学习,适用于打开办公室门的任务与移动机械手。使用额外的传感器输入(例如RGB +深度相机)来增强策略是提高机器人感知能力的直接方法,特别是对于在不同情况下可能有利于不同传感器的任务。随着我们将多传感器机器人学习扩展到非结构化的现实世界环境(例如办公室,家庭)和更复杂的机器人行为,我们也增加了对模拟器的依赖,以提高成本,效率和安全性。因此,跨多个传感器模态的模拟到真实的差距也增加了,使得模拟验证更加困难。我们表明,使用变分信息瓶颈(Alemi等人,2016)来正则化卷积神经网络,改进了对保留域的泛化,并以传感器不可知的方式减少了模拟到真实的差距。作为副作用,学习的嵌入还为每个传感器提供了模型不确定性的有用估计。我们证明,我们的方法是能够帮助关闭模拟到真实的差距,并成功地融合RGB和深度模态的基础上了解每个传感器的情况的不确定性。在真实的办公环境中,我们实现了96%的任务成功率,比基线提高了+16%。

英文摘要:

In this work we investigate and demonstrate benefits of a Bayesian approach to imitation learning from multiple sensor inputs, as applied to the task of opening office doors with a mobile manipulator. Augmenting policies with additional sensor inputs, such as RGB + depth cameras, is a straightforward approach to improving robot perception capabilities, especially for tasks that may favor different sensors in different situations. As we scale multi-sensor robotic learning to unstructured real-world settings (e.g. offices, homes) and more complex robot behaviors, we also increase reliance on simulators for cost, efficiency, and safety. Consequently, the sim-to-real gap across multiple sensor modalities also increases, making simulated validation more difficult. We show that using the Variational Information Bottleneck (Alemi et al., 2016) to regularize convolutional neural networks improves generalization to held-out domains and reduces the sim-to-real gap in a sensor-agnostic manner. As a side effect, the learned embeddings also provide useful estimates of model uncertainty for each sensor. We demonstrate that our method is able to help close the sim-to-real gap and successfully fuse RGB and depth modalities based on understanding of the situational uncertainty of each sensor. In a real-world office environment, we achieve 96% task success, improving upon the baseline by +16%.

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。

↑