Data efficiency and robustness to task-irrelevant perturbations are long-standing challenges for deep reinforcement learning algorithms. Here we introduce a modular approach to addressing these challenges in a continuous control environment, without using hand-crafted or supervised information. Our Curious Object-Based seaRch Agent (COBRA) uses task-free intrinsically motivated exploration and unsupervised learning to build object-based models of its environment and action space. Subsequently, it can learn a variety of tasks through model-based search in very few steps and excel on structured hold-out tests of policy robustness.
@article{arxiv.1905.09275,
title = {COBRA: Data-Efficient Model-Based RL through Unsupervised Object Discovery and Curiosity-Driven Exploration},
author = {Nicholas Watters and Loic Matthey and Matko Bosnjak and Christopher P. Burgess and Alexander Lerchner},
journal= {arXiv preprint arXiv:1905.09275},
year = {2019}
}