Learning robust driving policies without online exploration
Abstract
We propose a multi-time-scale predictive representation learning method to efficiently learn robust driving policies in an offline manner that generalize well to novel road geometries, and damaged and distracting lane conditions which are not covered in the offline training data. We show that our proposed representation learning method can be applied easily in an offline (batch) reinforcement learning setting demonstrating the ability to generalize well and efficiently under novel conditions compared to standard batch RL methods. Our proposed method utilizes training data collected entirely offline in the real-world which removes the need of intensive online explorations that impede applying deep reinforcement learning on real-world robot training. Various experiments were conducted in both simulator and real-world scenarios for the purpose of evaluation and analysis of our proposed claims.
Cite
@article{arxiv.2103.08070,
title = {Learning robust driving policies without online exploration},
author = {Daniel Graves and Nhat M. Nguyen and Kimia Hassanzadeh and Jun Jin and Jun Luo},
journal= {arXiv preprint arXiv:2103.08070},
year = {2021}
}
Comments
Accepted in ICRA 2021. Due to format limitations of ICRA, we include appendix of our detailed evaluation results in this full version. arXiv admin note: substantial text overlap with arXiv:2006.15110