Saved in:
Bibliographic Details
Main Authors: Qian, Kangan, Luo, Ziang, Jiang, Sicong, Huang, Zilin, Miao, Jinyu, Ma, Zhikun, Zhu, Tianze, Li, Jiayin, He, Yangfan, Fu, Zheng, Shi, Yining, Wang, Boyue, Lin, Hezhe, Chen, Ziyu, Yu, Jiangbo, Jiao, Xinyu, Yang, Mengmeng, Jiang, Kun, Yang, Diange
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.08162
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917951246958592
author Qian, Kangan
Luo, Ziang
Jiang, Sicong
Huang, Zilin
Miao, Jinyu
Ma, Zhikun
Zhu, Tianze
Li, Jiayin
He, Yangfan
Fu, Zheng
Shi, Yining
Wang, Boyue
Lin, Hezhe
Chen, Ziyu
Yu, Jiangbo
Jiao, Xinyu
Yang, Mengmeng
Jiang, Kun
Yang, Diange
author_facet Qian, Kangan
Luo, Ziang
Jiang, Sicong
Huang, Zilin
Miao, Jinyu
Ma, Zhikun
Zhu, Tianze
Li, Jiayin
He, Yangfan
Fu, Zheng
Shi, Yining
Wang, Boyue
Lin, Hezhe
Chen, Ziyu
Yu, Jiangbo
Jiao, Xinyu
Yang, Mengmeng
Jiang, Kun
Yang, Diange
contents Ensuring safe, comfortable, and efficient planning is crucial for autonomous driving systems. While end-to-end models trained on large datasets perform well in standard driving scenarios, they struggle with complex low-frequency events. Recent Large Language Models (LLMs) and Vision Language Models (VLMs) advancements offer enhanced reasoning but suffer from computational inefficiency. Inspired by the dual-process cognitive model "Thinking, Fast and Slow", we propose $\textbf{FASIONAD}$ -- a novel dual-system framework that synergizes a fast end-to-end planner with a VLM-based reasoning module. The fast system leverages end-to-end learning to achieve real-time trajectory generation in common scenarios, while the slow system activates through uncertainty estimation to perform contextual analysis and complex scenario resolution. Our architecture introduces three key innovations: (1) A dynamic switching mechanism enabling slow system intervention based on real-time uncertainty assessment; (2) An information bottleneck with high-level plan feedback that optimizes the slow system's guidance capability; (3) A bidirectional knowledge exchange where visual prompts enhance the slow system's reasoning while its feedback refines the fast planner's decision-making. To strengthen VLM reasoning, we develop a question-answering mechanism coupled with reward-instruct training strategy. In open-loop experiments, FASIONAD achieves a $6.7\%$ reduction in average $L2$ trajectory error and $28.1\%$ lower collision rate.
format Preprint
id arxiv_https___arxiv_org_abs_2503_08162
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FASIONAD++ : Integrating High-Level Instruction and Information Bottleneck in FAt-Slow fusION Systems for Enhanced Safety in Autonomous Driving with Adaptive Feedback
Qian, Kangan
Luo, Ziang
Jiang, Sicong
Huang, Zilin
Miao, Jinyu
Ma, Zhikun
Zhu, Tianze
Li, Jiayin
He, Yangfan
Fu, Zheng
Shi, Yining
Wang, Boyue
Lin, Hezhe
Chen, Ziyu
Yu, Jiangbo
Jiao, Xinyu
Yang, Mengmeng
Jiang, Kun
Yang, Diange
Robotics
Computation and Language
Ensuring safe, comfortable, and efficient planning is crucial for autonomous driving systems. While end-to-end models trained on large datasets perform well in standard driving scenarios, they struggle with complex low-frequency events. Recent Large Language Models (LLMs) and Vision Language Models (VLMs) advancements offer enhanced reasoning but suffer from computational inefficiency. Inspired by the dual-process cognitive model "Thinking, Fast and Slow", we propose $\textbf{FASIONAD}$ -- a novel dual-system framework that synergizes a fast end-to-end planner with a VLM-based reasoning module. The fast system leverages end-to-end learning to achieve real-time trajectory generation in common scenarios, while the slow system activates through uncertainty estimation to perform contextual analysis and complex scenario resolution. Our architecture introduces three key innovations: (1) A dynamic switching mechanism enabling slow system intervention based on real-time uncertainty assessment; (2) An information bottleneck with high-level plan feedback that optimizes the slow system's guidance capability; (3) A bidirectional knowledge exchange where visual prompts enhance the slow system's reasoning while its feedback refines the fast planner's decision-making. To strengthen VLM reasoning, we develop a question-answering mechanism coupled with reward-instruct training strategy. In open-loop experiments, FASIONAD achieves a $6.7\%$ reduction in average $L2$ trajectory error and $28.1\%$ lower collision rate.
title FASIONAD++ : Integrating High-Level Instruction and Information Bottleneck in FAt-Slow fusION Systems for Enhanced Safety in Autonomous Driving with Adaptive Feedback
topic Robotics
Computation and Language
url https://arxiv.org/abs/2503.08162