提出FluidTest评估框架，检测自动驾驶规划中的隐藏安全风险。
Is Your Trajectory Displacement Safe in Long-tail?
- 将评估转化为威胁检测，通过人类标注与三智能体验证结合
- 在WOD-E2E数据集上发现65%的Poutine轨迹存在新增威胁
- 适合关注自动驾驶安全评估的开发者与研究者
长尾场景仍是自动驾驶评估的主要瓶颈，尽管数据集规模已大幅增长。现有评估流程通常缺乏与人类一致、安全敏感、可验证和可解释性：闭环指标常在优秀规划器间饱和，而无结构的人类评分则可能因设计不当产生噪声。本文将规划评估定义为额外威胁检测：给定规划轨迹与专家参考轨迹，该规划的位移是否引入新的不安全驾驶行为？提出FluidTest评估流水线，包含三项组件：用于可靠人工标注的成对WebUI协议；基于证据的32类语义威胁分类体系与决策图；具备反思能力的三智能体验证系统以保证精度与可审计性。在WOD-E2E数据集上的实验表明，FluidTest能实现训练标注者间的一致标签，且识别出65%的Poutine轨迹和51%的RAP轨迹中存在额外威胁。结果表明，即使具有高评分反馈分（RFS）和低平均位移误差（ADE），当前最先进规划器仍存在显著安全相关缺陷。更多细节、指南与代码见https://fluidtest.web.app。
原文摘要 · Abstract (English)
Long-tail scenarios remain a major bottleneck for autonomous driving evaluation, even as datasets grow by orders of magnitude. Existing evaluation pipelines are rarely human-aligned, safety-aware, verifiable, and explainable at the same time: closed-loop metrics often saturate among strong planners, while unstructured human ratings can be noisy without a carefully designed protocol. We formulate planning evaluation as additional-threat detection: given a planner trajectory and an expert reference, does the planner's displacement introduce new unsafe driving behavior? We propose FluidTest, an evaluation pipeline with three components: a pairwise WebUI protocol for reliable human annotation; a taxonomy of 32 semantic threats with evidence-grounded decision graphs; and a three-agent verification system with reflection for precision and auditability. Experiments on the WOD-E2E dataset show that FluidTest produces consistent labels among trained annotators and identifies additional threats in 65% of Poutine trajectories and 51% of RAP trajectories. These results show that state-of-the-art planners can still exhibit substantial safety-relevant failures despite high Rater Feedback Scores (RFS) and low Average Displacement Error (ADE). Additional details, guidance, and code are available at https://fluidtest.web.app.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。