DriveMind：基于信念演化的闭环自动驾驶框架
DriveMind: Mind-Evolving Belief Tracking for Closed-Loop Autonomous Driving
AI 导读
研究者提出 DriveMind，一个基于多模态大语言模型的自动驾驶框架，借鉴粒子滤波思想同时维护多个场景假设，并随新帧到来更新权重。在 Bench2Drive 上，加入信念追踪后 Driving Score 从 68.50 提升至 81.37，Success Rate 从 41.36% 提升至 54.09%，在无信号路口和会车交互等模糊场景上提升最大（+20.0 / +20.5 DS）。
正文 · 原文
Abstract:Recent multimodal large language models (MLLMs) can perceive a traffic scene, reason about it, and propose a driving action within a single model. However, existing MLLM-based drivers reason from scratch at each frame and commit to a single explanation, discarding earlier hypotheses about the scene. When the scene is ambiguous, the agent locks in on the wrong interpretation and acts on it before evidence accumulates. In this paper, we propose DriveMind, an MLLM-based driving framework that keeps several hypotheses about the scene alive at the same time and updates their weights as new frames arrive. Inspired by particle filtering, DriveMind propagates each hypothesis through a VLM, scores it against the current observation, and resamples to maintain diversity. On Bench2Drive, adding belief tracking to the same driving pipeline improves the Driving Score (DS) from 68.50 to 81.37 and the Success Rate (SR) from 41.36% to 54.09%. The largest gains occur on ambiguous scenarios such as unsignalized intersections (+20.0 DS) and oncoming-traffic interactions (+20.5 DS). Beyond the overall metrics, DriveMind also improves fine-grained interactive driving abilities such as merging, emergency braking, and give-way.
| Comments: | Accepted by Neurips 2026 |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.32069 [cs.AI] |
| (or arXiv:2610.32069v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.32069 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yan Shi [view email]
[v1]
Fri, 23 Oct 2026 02:13:50 UTC (1,331 KB)
来源：arXiv cs.AI · arxiv.org