2018-10-23-golemo18a.md

title	abstract	layout	series	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	publisher	container-title	volume	genre	issued	pdf	extras
Sim-to-Real Transfer with Neural-Augmented Robot Simulation	Despite the recent successes of deep reinforcement learning, teaching complex motor skills to a physical robot remains a hard problem. While learning directly on a real system is usually impractical, doing so in simulation has proven to be fast and safe. Nevertheless, because of the "reality gap," policies trained in simulation often perform poorly when deployed on a real system. In this work, we introduce a method for training a recurrent neural network on the differences between simulated and real robot trajectories and then using this model to augment the simulator. This Neural-Augmented Simulation (NAS) can be used to learn control policies that transfer significantly better to real environments than policies learned on existing simulators. We demonstrate the potential of our approach through a set of experiments on the Mujoco simulator with added backlash and the Poppy Ergo Jr robot. NAS allows us to learn policies that are competitive with ones that would have been learned directly on the real robot.	inproceedings	Proceedings of Machine Learning Research	golemo18a	0	Sim-to-Real Transfer with Neural-Augmented Robot Simulation	817	828	817-828	817	false	Golemo, Florian and Taiga, Adrien Ali and Courville, Aaron and Oudeyer, Pierre-Yves	given family Florian Golemo given family Adrien Ali Taiga given family Aaron Courville given family Pierre-Yves Oudeyer	2018-10-23		PMLR	Proceedings of The 2nd Conference on Robot Learning	87	inproceedings	date-parts 2018 10 23	http://proceedings.mlr.press/v87/golemo18a/golemo18a.pdf	label link Supplementary video https://youtu.be/nw9YTbCEYH8 label link Source code https://github.com/aalitaiga/sim-to-real/