Simulation-based reinforcement learning for real-world autonomous driving
Machine Learning
2024-04-04 v4 Artificial Intelligence
Robotics
Abstract
We use reinforcement learning in simulation to obtain a driving system controlling a full-size real-world vehicle. The driving policy takes RGB images from a single camera and their semantic segmentation as input. We use mostly synthetic data, with labelled real-world data appearing only in the training of the segmentation network. Using reinforcement learning in simulation and synthetic data is motivated by lowering costs and engineering effort. In real-world experiments we confirm that we achieved successful sim-to-real policy transfer. Based on the extensive evaluation, we analyze how design decisions about perception, control, and training impact the real-world performance.
Cite
@article{arxiv.1911.12905,
title = {Simulation-based reinforcement learning for real-world autonomous driving},
author = {Błażej Osiński and Adam Jakubowski and Piotr Miłoś and Paweł Zięcina and Christopher Galias and Silviu Homoceanu and Henryk Michalewski},
journal= {arXiv preprint arXiv:1911.12905},
year = {2024}
}