Saved in:
Bibliographic Details
Main Authors: Tang, Zecong, Wang, Zixu, Wang, Yifei, Lian, Weitong, Gao, Tianjian, Li, Haoran, Ru, Tengju, Meng, Lingyi, Cui, Zhejun, Zhu, Yichen, Kang, Qi, Wang, Kaixuan, Zhang, Yu
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.14702
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917535440437248
author Tang, Zecong
Wang, Zixu
Wang, Yifei
Lian, Weitong
Gao, Tianjian
Li, Haoran
Ru, Tengju
Meng, Lingyi
Cui, Zhejun
Zhu, Yichen
Kang, Qi
Wang, Kaixuan
Zhang, Yu
author_facet Tang, Zecong
Wang, Zixu
Wang, Yifei
Lian, Weitong
Gao, Tianjian
Li, Haoran
Ru, Tengju
Meng, Lingyi
Cui, Zhejun
Zhu, Yichen
Kang, Qi
Wang, Kaixuan
Zhang, Yu
contents Autonomous driving requires reliable perception and safe decision-making in complex scenarios. Recent vision-language models (VLMs) demonstrate reasoning and generalization abilities, opening new possibilities for autonomous driving; however, existing benchmarks often evaluate perception and decision-making separately, limit failure analysis with choice-only formats, or introduce evaluation bias through LLM-scored long-form outputs. To address these issues, we present Drive-P2D, a progressive perception-to-decision benchmark with 6,650 questions across Object, Scene, and Decision levels. Drive-P2D adopts a separated reasoning-and-answer protocol: final answers are scored objectively, while reasoning is analyzed to identify error modes exposed along the progressive perception-to-decision chain. We evaluate mainstream VLMs across all and high-risk scenarios, and further characterize the perception-to-decision capability boundary through correlation analysis and similar-scene robustness testing. Reasoning further exposes failure modes such as logical reasoning errors and semantic feature omissions, and we train a lightweight analyzer model to automate large-scale error-mode annotation of reasoning. Together, these designs provide practical insights for building safer and more reliable VLMs for real-world autonomous driving.
format Preprint
id arxiv_https___arxiv_org_abs_2601_14702
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Drive-P2D: A Progressive Perception-to-Decision Benchmark for VLMs in Autonomous Driving
Tang, Zecong
Wang, Zixu
Wang, Yifei
Lian, Weitong
Gao, Tianjian
Li, Haoran
Ru, Tengju
Meng, Lingyi
Cui, Zhejun
Zhu, Yichen
Kang, Qi
Wang, Kaixuan
Zhang, Yu
Artificial Intelligence
Computer Vision and Pattern Recognition
Robotics
Autonomous driving requires reliable perception and safe decision-making in complex scenarios. Recent vision-language models (VLMs) demonstrate reasoning and generalization abilities, opening new possibilities for autonomous driving; however, existing benchmarks often evaluate perception and decision-making separately, limit failure analysis with choice-only formats, or introduce evaluation bias through LLM-scored long-form outputs. To address these issues, we present Drive-P2D, a progressive perception-to-decision benchmark with 6,650 questions across Object, Scene, and Decision levels. Drive-P2D adopts a separated reasoning-and-answer protocol: final answers are scored objectively, while reasoning is analyzed to identify error modes exposed along the progressive perception-to-decision chain. We evaluate mainstream VLMs across all and high-risk scenarios, and further characterize the perception-to-decision capability boundary through correlation analysis and similar-scene robustness testing. Reasoning further exposes failure modes such as logical reasoning errors and semantic feature omissions, and we train a lightweight analyzer model to automate large-scale error-mode annotation of reasoning. Together, these designs provide practical insights for building safer and more reliable VLMs for real-world autonomous driving.
title Drive-P2D: A Progressive Perception-to-Decision Benchmark for VLMs in Autonomous Driving
topic Artificial Intelligence
Computer Vision and Pattern Recognition
Robotics
url https://arxiv.org/abs/2601.14702