Simulation-Based Reinforcement Learning for Real-World Autonomous Driving
2020pp. 6411–6418
Citations Over TimeTop 19% of 2020 papers
Błażej Osiński, Adam Jakubowski, Pawel Ziecina, Piotr Milos, Christopher Galias, Silviu Homoceanu, Henryk Michalewski
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.
Related Papers
- → Real–Sim–Real Transfer for Real-World Robot Control Policy Learning with Deep Reinforcement Learning(2020)28 cited
- → Fast Prototype Framework for Deep Reinforcement Learning-based Trajectory Planner(2020)6 cited
- → Effects of Domain Randomization on Simulation-to-Reality Transfer of Reinforcement Learning Policies for Industrial Robots(2021)3 cited
- → Virtual to Real Reinforcement Learning for Autonomous Driving(2017)149 cited