• Classified by Topic • Classified by Publication Type • Sorted by Date • Sorted by First Author Last Name • Classified by Funding Source •
A Champion-Level Vision-Based Reinforcement Learning Agent for Competitive Racing in Gran Turismo 7.
Hojoon
Lee, Takuma Seno, Jun Jet Tai, Kaushik Subramanian, Kenta Kawamoto, Peter Stone,
and Peter R. Wurman.
IEEE Robotics and Automation Letters, 10(6):5545–52,
June 2025.
2.5-minute Video Summary
Deep reinforcement learning has achieved superhuman racing performance in high-fidelity simulators like Gran Turismo 7 (GT7). It typically utilizes global features that require instrumentation external to a car, such as precise localization of agents and opponents, limiting real-world applicability. To address this limitation, we introduce a vision-based autonomous racing agent that relies solely on ego-centric camera views and onboard sensor data, eliminating the need for precise localization during inference. This agent employs an asymmetric actor-critic framework: the actor uses a recurrent neural network with the sensor data local to the car to retain track layouts and opponent positions, while the critic accesses the global features during training. Evaluated in GT7, our agent consistently outperforms GT7's built-drivers. To our knowledge, this work presents the first vision-based autonomous racing agent to demonstrate champion-level performance in competitive racing scenarios.
@Article{peter_stone_ral_2025,
author = {Hojoon Lee and Takuma Seno and Jun Jet Tai and Kaushik Subramanian and Kenta Kawamoto and Peter Stone and Peter R.\ Wurman},
title = {A Champion-Level Vision-Based Reinforcement Learning Agent for Competitive Racing in {G}ran {T}urismo 7},
journal = {{IEEE} Robotics and Automation Letters},
month="June",
Volume="10",Number="6",
pages="5545--52",
doi="10.1109/LRA.2025.3560873",
year = {2025},
abstract = {Deep reinforcement learning has achieved superhuman racing performance in high-fidelity simulators like Gran Turismo 7 (GT7). It typically utilizes global features that require instrumentation external to a car, such as precise localization of agents and opponents, limiting real-world applicability. To address this limitation, we introduce a vision-based autonomous racing agent that relies solely on ego-centric camera views and onboard sensor data, eliminating the need for precise localization during inference. This agent employs an asymmetric actor-critic framework: the actor uses a recurrent neural network with the sensor data local to the car to retain track layouts and opponent positions, while the critic accesses the global features during training. Evaluated in GT7, our agent consistently outperforms GT7's built-drivers. To our knowledge, this work presents the first vision-based autonomous racing agent to demonstrate champion-level performance in competitive racing scenarios.},
wwwnote={<a href="https://www.youtube.com/watch?v=cWqKbFsJcpo">2.5-minute Video Summary</a>},
}
Generated by bib2html.pl (written by Patrick Riley ) on Wed Jun 10, 2026 15:26:41