We present a framework for data-driven robotics that makes use of a large dataset of recorded robot experience and scales to several tasks using learned reward functions. We show how to apply this framework to accomplish three different object manipulation tasks on a real robot platform. Given demonstrations of a task together with task-agnostic recorded experience, we use a special form of human annotation as supervision to learn a reward function, which enables us to deal with real-world tasks where the reward signal cannot be acquired directly. Learned rewards are used in combination with a large dataset of experience from different tasks to learn a robot policy offline using batch RL. We show that using our approach it is possible to train agents to perform a variety of challenging manipulation tasks including stacking rigid objects and handling cloth.
@article{arxiv.1909.12200,
title = {Scaling data-driven robotics with reward sketching and batch reinforcement learning},
author = {Serkan Cabi and Sergio Gómez Colmenarejo and Alexander Novikov and Ksenia Konyushkova and Scott Reed and Rae Jeong and Konrad Zolna and Yusuf Aytar and David Budden and Mel Vecerik and Oleg Sushkov and David Barker and Jonathan Scholz and Misha Denil and Nando de Freitas and Ziyu Wang},
journal= {arXiv preprint arXiv:1909.12200},
year = {2020}
}