Summary of Haim-drl: Enhanced Human-in-the-loop Reinforcement Learning For Safe and Efficient Autonomous Driving, by Zilin Huang et al.
HAIM-DRL: Enhanced Human-in-the-loop Reinforcement Learning for Safe and Efficient Autonomous Driving
by Zilin Huang, Zihao Sheng, Chengyuan Ma, Sikai Chen
First submitted to arxiv on: 6 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); 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 abstract proposes a novel method, Human as AI mentor-based deep reinforcement learning (HAIM-DRL), for safe and efficient autonomous driving in mixed traffic platoons. The approach leverages human intelligence by introducing an innovative learning paradigm, Human as AI mentor (HAIM). This paradigm allows the AI agent to explore uncertain environments while a human expert takes control in dangerous situations, providing correct actions to avoid accidents. HAIM-DRL uses data from free exploration and partial human demonstrations to train the agent, eliminating the need for manually designing reward functions. The method also employs minimal intervention techniques to reduce cognitive load on the human mentor. Comparative results show that HAIM-DRL outperforms traditional methods in driving safety, sampling efficiency, traffic flow disturbance mitigation, and generalizability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper proposes a new way for self-driving cars to work with humans. It’s called Human as AI mentor-based deep reinforcement learning (HAIM-DRL). This method lets the car learn from both its own exploration of the road and guidance from a human expert. The human can take control when things get tricky, showing the car what to do. This helps keep everyone safe on the road. The new approach is better than other methods at keeping traffic flowing smoothly. |
Keywords
* Artificial intelligence * Reinforcement learning
