Research map / Vision-language and multimodal / Multimodal LLMs
DriveMind: Mind-Evolving Belief Tracking for Closed-Loop Autonomous Driving
Yan Shi, Kehua Chen, Jiarui Qi, Yihong Tang, Yinhai Wang, Liao Haicheng
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes DriveMind, a mind-evolving belief-tracking approach for MLLM-based closed-loop autonomous driving that maintains and revises hypotheses about the scene across frames instead of reasoning from scratch each frame.
Problem
MLLM-based autonomous drivers reason from scratch at each frame and commit to a single scene explanation, discarding earlier hypotheses, so under ambiguity they lock onto a wrong interpretation and act before evidence accumulates.
Approach
Maintains an evolving set of belief hypotheses about the traffic scene that is updated as evidence accumulates over frames, rather than committing to a single per-frame explanation. Full method details are not available in the provided text, so this is inferred from the abstract only.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add multi-hypothesis belief tracking (K=4) to the same VLM and diffusion controller (ablation) | DriveMind · Bench2Drive | Driving Score: 68.50 -> 81.37 | Improved |
| add multi-hypothesis belief tracking (K=4) (ablation) | DriveMind · Bench2Drive | Success Rate (%): 41.36 -> 54.09 | Improved |
| use 4 hypotheses instead of 2 (also adds deceleration and lane-change initial hypotheses) (ablation) | DriveMind belief tracking · Bench2Drive | Driving Score: 70.00 -> 81.37 | Improved |
| weight hypotheses by VLM consistency score instead of uniform weights (ablation) | DriveMind belief weighting · Bench2Drive maneuver prediction | Maneuver accuracy: 0.779 -> 0.930 | Improved |
| weight hypotheses by VLM score instead of random weights (ablation) | DriveMind belief weighting · Bench2Drive maneuver prediction | Maneuver accuracy: 0.644 -> 0.930 | Improved |
| increase hypothesis count K beyond 4 (ablation) | DriveMind belief tracking · behavioral diversity (four scenarios) | action disagreement / unique predictions | No clear effect |
| use asynchronous slow-fast belief refresh instead of synchronous reasoning (ablation) | DriveMind · Bench2Drive runtime (RTX A6000) | control-step latency (ms): 177.0 -> 65.5 | Improved |
| increase hypothesis count K from 2 to 6 (ablation) | High-level belief update · Bench2Drive runtime | latency (ms): 74.0 -> 165.5 | Hurt |
| add belief tracking (ablation) | DriveMind · NAVSIM | PDMS: 78.2 -> 81.0 | Improved |
| add belief tracking | DriveMind · Bench2Drive Signalized Intersection | Success Rate (%): 12.0 -> 4.0 | Hurt |
| add belief tracking | DriveMind · Bench2Drive Opposite Direction | Success Rate (%): 13.3 -> 13.3 | No clear effect |
| add belief tracking | DriveMind · Bench2Drive Unsignalized Intersection | Driving Score: 54.4 -> 74.4 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ReCogDrive — ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving (ICLR 2026) · better on 3 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Limitations
- Full text unavailable; method and results cannot be verified from the abstract alone.
Closest papers by meaning
- DriveCombo: Benchmarking Compositional Traffic Rule Reasoning in Autonomous Driving (CVPR 2026 · similarity 0.92)
- DriveSpatial: A Benchmark for Spatiotemporal Intelligence in VLMs for Autonomous Driving (NeurIPS 2026 · similarity 0.90)
- MindDriver: Introducing Progressive Multimodal Reasoning for Autonomous Driving (CVPR 2026 · similarity 0.89)
- SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models (ICML 2025 · similarity 0.89)
- nuReasoning: A Reasoning-Centric Dataset and Benchmark for Long-Tail Autonomous Driving (NeurIPS 2026 · similarity 0.89)
- DriveAgent-R1: Advancing VLM-based Autonomous Driving with Active Perception and Hybrid Thinking (ICLR 2026 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
