Learning to control robots directly based on images is a primary challenge in robotics. However, many existing reinforcement learning approaches require iteratively obtaining millions of robot samples to learn a policy, which can take significant time. In this paper, we focus on learning a realistic world model capturing the dynamics of scene changes conditioned on robot actions. Our dreaming model can emulate samples equivalent to a sequence of images from the actual environment, technically by learning an action-conditioned future representation/scene regressor. This allows the agent to learn action policies (i.e., visuomotor policies) by interacting with the dreaming model rather than the real-world. We experimentally confirm that our dreaming model enables robot learning of policies that transfer to the real-world.
@article{arxiv.1805.07813,
title = {Learning Real-World Robot Policies by Dreaming},
author = {AJ Piergiovanni and Alan Wu and Michael S. Ryoo},
journal= {arXiv preprint arXiv:1805.07813},
year = {2019}
}