Summary of A Super-human Vision-based Reinforcement Learning Agent For Autonomous Racing in Gran Turismo, by Miguel Vasco et al.
A Super-human Vision-based Reinforcement Learning Agent for Autonomous Racing in Gran Turismo
by Miguel Vasco, Takuma Seno, Kenta Kawamoto, Kaushik Subramanian, Peter R. Wurman, Peter Stone
First submitted to arxiv on: 18 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents an end-to-end deep reinforcement learning agent that can outperform human drivers in a high-fidelity racing simulator, Gran Turismo, using only local sensor data from an ego-centric camera view and on-board quantities such as velocity. The agent leverages global features for training but relies solely on local inputs during actual racing, achieving super-human performance. By introducing the first vision-based super-human car racing agent, this paper showcases the potential of AI in autonomous vehicles. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The researchers developed an AI system that can drive a car faster than humans by using only cameras and sensors inside the car. The system was trained to use global features, but it uses its own camera view and speed data to make decisions during racing. This is the first time a system has been able to beat human drivers in a racing simulator using just local data. |
Keywords
» Artificial intelligence » Reinforcement learning
