Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-to-End Learning from Demonstration
Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-to-End Learning from Demonstration
复制标题
DOI：
10.1109/icra.2018.8461076
复制
发表时间：
2017-07
期刊：
影响因子：
--
通讯作者：
Rouhollah Rahmatizadeh;P. Abolghasemi;Ladislau Bölöni;S. Levine
中科院分区：
文献类型：
--
作者：
Rouhollah Rahmatizadeh;P. Abolghasemi;Ladislau Bölöni;S. Levine
我们提出了一种多任务学习技术，该技术训练低成本机械臂的控制器完成一些复杂的拾取和放置任务，以及非握握性操作。控制器是一个循环神经网络，使用原始图像作为输入，生成机器人手臂轨迹，并在各个任务之间共享参数。该控制器还将基于vae - gan的重构与自回归多模态动作预测相结合。我们的研究结果表明，完全从具有直接行为克隆的原始图像中学习复杂的操作任务是可能的，例如拿起毛巾，擦拭物体，并将毛巾放置到先前的位置。我们发现权值共享和基于重构的正则化极大地提高了泛化和鲁棒性，并且在多个任务上同时训练提高了所有任务的成功率。
We propose a technique for multi-task learning from demonstration that trains the controller of a low-cost robotic arm to accomplish several complex picking and placing tasks, as well as non-prehensile manipulation. The controller is a recurrent neural network using raw images as input and generating robot arm trajectories, with the parameters shared across the tasks. The controller also combines VAE-GAN-based reconstruction with autoregressive multimodal action prediction. Our results demonstrate that it is possible to learn complex manipulation tasks, such as picking up a towel, wiping an object, and depositing the towel to its previous position, entirely from raw images with direct behavior cloning. We show that weight sharing and reconstruction-based regularization substantially improve generalization and robustness, and training on multiple tasks simultaneously increases the success rate on all tasks.