DriveAction: A Benchmark for Exploring Human-like Driving Decisions in VLA Models
DriveAction: A Benchmark for Exploring Human-like Driving Decisions in VLA Models
Vision-Language-Action (VLA) models have advanced autonomous driving, but existing benchmarks still lack scenario diversity, reliable action-level annotation, and evaluation protocols aligned with human preferences. To address these limitations, we introduce DriveAction, the first action-driven benchmark specifically designed for VLA models, comprising 16,185 QA pairs generated from 2,610 driving scenarios. DriveAction leverages real-world driving data proactively collected by users of production-level autonomous vehicles to ensure broad and representative scenario coverage, offers high-level discrete action labels collected directly from users' actual driving operations, and implements an action-rooted tree-structured evaluation framework that explicitly links vision, language, and action tasks, supporting both comprehensive and task-specific assessment. Our experiments demonstrate that state-of-the-art vision-language models (VLMs) require both vision and language guidance for accurate action prediction: on average, accuracy drops by 3.3% without vision input, by 4.1% without language input, and by 8.0% without either. Our evaluation supports precise identification of model bottlenecks with robust and consistent results, thus providing new insights and a rigorous foundation for advancing human-like decisions in autonomous driving.
Yuhan Hao、Zhengning Li、Lei Sun、Weilong Wang、Naixin Yi、Sheng Song、Caihong Qin、Mofan Zhou、Yifei Zhan、Peng Jia、Xianpeng Lang
自动化技术、自动化技术设备计算技术、计算机技术
Yuhan Hao,Zhengning Li,Lei Sun,Weilong Wang,Naixin Yi,Sheng Song,Caihong Qin,Mofan Zhou,Yifei Zhan,Peng Jia,Xianpeng Lang.DriveAction: A Benchmark for Exploring Human-like Driving Decisions in VLA Models[EB/OL].(2025-06-05)[2025-06-25].https://arxiv.org/abs/2506.05667.点此复制
评论