ImagiDrive: A Unified Imagination-and-Planning Framework for Autonomous Driving

arXiv 25.8 2025 70.6 method, system, application

TLDR

ImagiDrive integrates a VLM driving agent with a DWM scene imaginer for iterative planning in autonomous driving.

Reasoning

The paper proposes a novel framework combining vision-language models and driving world models for autonomous driving, addressing integration challenges. Strengths include a clear motivation and a unified loop; weaknesses are that the abstract is cut off, lacking explicit results and real-world validation.

Read-first score

Read-first score 70.6, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 43.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=True; dataset=False; markers=dataset,github

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=dataset,experiment,validation

Topical relevance 42%

61.4

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Field roles

FrontierMethodology anchorReproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 175.

Keyword Scores

world model

generative world model

world dynamics prediction

interactive world model

world simulator

video world model

model-based reinforcement learning world model

Deep Analysis

Innovations

Unified imagination-and-planning loop integrating a VLM-based driving agent with a DWM-based scene imaginer
Early stopping mechanism to improve computational efficiency
Trajectory selection strategy to enhance predictive accuracy

Methodology

ImagiDrive is an end-to-end autonomous driving framework that combines a VLM-based driving agent with a DWM-based scene imaginer. The agent first predicts initial driving trajectories from multi-modal inputs, then the imaginer generates corresponding future scenarios, which are iteratively used to refine the agent's planning decisions. An early stopping mechanism and a trajectory selection strategy are introduced to address efficiency and accuracy challenges. The model is evaluated on the nuScenes and NAVSIM datasets under both open-loop and closed-loop conditions.

Key Results

Extensive experiments on nuScenes and NAVSIM demonstrate that ImagiDrive achieves robustness and superiority over previous alternatives in both open-loop and closed-loop driving scenarios.

Resources

Paper Code