Reproduce “A Survey on Large Language Model-empowered Autonomous Driving” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yuxuan Zhu, Shiyi Wang, Wenqing Zhong, Nianchen Shen, Yunqi Li, Siqi Wang, Zhiheng Li, Cathy Wu and 2 more
A comprehensive survey analyzes the application of Large Language Models (LLMs) in autonomous driving (AD), evaluating their potential to address challenges across modular and end-to-end architectures. The paper identifies LLMs' advantages in common sense, reasoning, and communication, suggesting a move towards hybrid data and knowledge-driven AD systems, while also outlining current limitations.
Key insights
- LLMs bring human-like common sense, reasoning, and communication abilities that are crucial for understanding complex traffic scenarios and enhancing human-machine interaction in AD.
- The integration of LLMs is fostering a paradigm shift in AD research, moving from purely data-driven models to hybrid systems that leverage both data and knowledge.
- LLMs show promise in blurring the lines between modular and end-to-end AD, by enabling unified information channels and improving aspects like data annotation, scenario generation, and interpretable decision-making.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.