Back to Search
Start Over
Super-Human Performance in Gran Turismo Sport Using Deep Reinforcement Learning
- Publication Year :
- 2022
-
Abstract
- Autonomous car racing is a major challenge in robotics. It raises fundamental problems for classical approaches such as planning minimum-time trajectories under uncertain dynamics and controlling the car at the limits of its handling. Besides, the requirement of minimizing the lap time, which is a sparse objective, and the difficulty of collecting training data from human experts have also hindered researchers from directly applying learning-based approaches to solve the problem. In the present work, we propose a learning-based system for autonomous car racing by leveraging a high-fidelity physical car simulation, a course-progress proxy reward, and deep reinforcement learning. We deploy our system in Gran Turismo Sport, a world-leading car simulator known for its realistic physics simulation of different race cars and tracks, which is even used to recruit human race car drivers. Our trained policy achieves autonomous racing performance that goes beyond what had been achieved so far by the built-in AI, and, at the same time, outperforms the fastest driver in a dataset of over 50,000 human players.<br />Comment: Accepted for Publication at the IEEE Robotics and Automation Letters (RA-L) 2021, and International Conference on Robots and Automation (ICRA) 2021
- Subjects :
- FOS: Computer and information sciences
0209 industrial biotechnology
Computer Science - Machine Learning
2606 Control and Optimization
Control and Optimization
1707 Computer Vision and Pattern Recognition
Computer science
Computer Science - Artificial Intelligence
10009 Department of Informatics
Biomedical Engineering
2210 Mechanical Engineering
2207 Control and Systems Engineering
2204 Biomedical Engineering
1702 Artificial Intelligence
02 engineering and technology
000 Computer science, knowledge & systems
Machine learning
computer.software_genre
Machine Learning (cs.LG)
Vehicle dynamics
Computer Science - Robotics
1709 Human-Computer Interaction
020901 industrial engineering & automation
Artificial Intelligence
0202 electrical engineering, electronic engineering, information engineering
1706 Computer Science Applications
Reinforcement learning
Training set
Artificial neural network
business.industry
Mechanical Engineering
Robotics
Computer Science Applications
Human-Computer Interaction
Artificial Intelligence (cs.AI)
Control and Systems Engineering
Dynamics (music)
Trajectory
020201 artificial intelligence & image processing
Computer Vision and Pattern Recognition
Dynamical simulation
Artificial intelligence
business
computer
Robotics (cs.RO)
Subjects
Details
- Language :
- English
- Database :
- OpenAIRE
- Accession number :
- edsair.doi.dedup.....1869fc371712794179e2abb615ff3851