Driving with A Thousand Faces: A Benchmark for Closed-Loop Personalized End-to-End Autonomous Driving

TL;DR

Person2Drive is proposed, a comprehensive personalized E2E-AD platform and benchmark that includes an open-source, flexible data collection system that simulates realistic scenarios to generate scalable and diverse personalized driving datasets; style vector-based evaluation metrics with Maximum Mean Discrepancy and KL divergence to comprehensively quantify individual driving behaviors.

Abstract

Human driving behavior is inherently diverse, yet most end-to-end autonomous driving (E2E-AD) systems learn a single average driving style, neglecting individual differences. Achieving personalized E2E-AD faces challenges across three levels: limited real-world datasets with individual-level annotations, a lack of quantitative metrics for evaluating personal driving styles, and the absence of algorithms that can learn stylized representations from users' trajectories. To address these gaps, we propose Person2Drive, a comprehensive personalized E2E-AD platform and benchmark. It includes an open-source, flexible data collection system that simulates realistic scenarios to generate scalable and diverse personalized driving datasets; style vector-based evaluation metrics with Maximum Mean Discrepancy and KL divergence to comprehensively quantify individual driving behaviors; and a personalized E2E-AD framework with a style reward model that efficiently adapts E2E models for safe and individualized driving. Extensive experiments demonstrate that Person2Drive enables fine-grained analysis, reproducible evaluation, and effective personalization in end-to-end autonomous driving. Our dataset and code will be released after acceptance.

Figures (10)

• Figure 1: Person2Drive framework. We develop a low-cost, scalable platform for data collection, featuring a navigation map and multi-camera views that mimic real-world driving habits, enabling authentic data acquisition across diverse scenarios, road segments, and driver styles. To achieve personalized driving style adaptation, we propose a three-stage adaptation framework. Finally, the adapted end-to-end model is deployed in a simulated environment for inference, and the style model is used to evaluate the effectiveness of personalization.
• Figure 2: Effect of style evaluation on the GT trajectories of the StyleDrive dataset, which reports average results among all styles.
• Figure 3: Style Modeling results on the Person2Drive dataset.
• Figure 4: Average style indices differences among all individuals.
• Figure 5: Comparison between baseline,gt and fine-tuned, showing that fine-tuning helps capture personalized driving styles.