论文标题

SEIL:模拟的模仿模仿学习

SEIL: Simulation-augmented Equivariant Imitation Learning

论文作者

Jia, Mingxi, Wang, Dian, Su, Guanang, Klee, David, Zhu, Xupeng, Walters, Robin, Platt, Robert

论文摘要

在机器人的操作中,获取样品非常昂贵,因为它通常需要与现实世界进行互动。传统的图像级数据增强表明有可能提高各种机器学习任务的样本效率。但是,图像级数据的增强不足以使模仿学习代理以合理的示范学习良好的操纵策略。我们提出了模拟的模拟模仿模仿学习(SEIL),该方法结合了一种新型的数据增强策略,该策略将专家轨迹与模拟过渡补充了,并利用了$ \ mathrm {o}(O}(2)$对称性的机器人操作中的$ \ mathrm {o}(2)。实验评估表明,我们的方法可以在十个演示中学习非平凡的操纵任务,并以显着的边缘优于基准。

In robotic manipulation, acquiring samples is extremely expensive because it often requires interacting with the real world. Traditional image-level data augmentation has shown the potential to improve sample efficiency in various machine learning tasks. However, image-level data augmentation is insufficient for an imitation learning agent to learn good manipulation policies in a reasonable amount of demonstrations. We propose Simulation-augmented Equivariant Imitation Learning (SEIL), a method that combines a novel data augmentation strategy of supplementing expert trajectories with simulated transitions and an equivariant model that exploits the $\mathrm{O}(2)$ symmetry in robotic manipulation. Experimental evaluations demonstrate that our method can learn non-trivial manipulation tasks within ten demonstrations and outperforms the baselines with a significant margin.

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源