arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.17586cs.ROcs.AIcs.LG
收藏

基于流匹配的连续异常检测:面向自动驾驶的流形感知频谱空间

Conditional Flow Matching for Continuous Anomaly Detection in Autonomous Driving on a Manifold-Aware Spectral Space

Independent Researcher(独立研究者)

Antonio Guillen-Perez

AI总结:

Deep-Flow通过流匹配技术在频谱空间中实现连续异常检测,利用运动学平滑性和低秩流形约束,提升自动驾驶安全验证的准确性与可靠性。

摘要:

在Level 4自动驾驶车辆的安全验证中,传统基于规则的启发式方法在检测罕见的高风险长尾场景时存在扩展瓶颈。我们提出了Deep-Flow,一种无监督的安全关键异常检测框架,利用最优传输条件流匹配(OT-CFM)来表征专家人类驾驶行为的连续概率密度。与标准生成方法在不稳定、高维坐标空间中操作不同,Deep-Flow通过主成分分析(PCA)瓶颈将生成过程限制到低秩频谱流形中。这通过设计确保运动学平滑性,并使计算精确的雅可比迹以实现数值稳定、确定性的对数似然估计。为了在复杂交叉口解决多模态模糊性,我们利用早期融合变换器编码器,具有车道感知的目标条件,其特征是直接连接到流头的跳过连接,以在整个网络中保持意图完整性。我们引入了一种运动学复杂性加权方案,在模拟自由训练过程中优先考虑高能量操作(通过路径曲折度和加速度量化)。在Waymo开放运动数据集(WOMD)上评估,我们的框架在与启发式黄金集安全关键事件的比较中实现了AUC-ROC 0.766。更显著的是,我们的分析揭示了运动学危险与语义不合规之间的根本区别。Deep-Flow通过揭示出分布外行为,如车道边界违规和非规范交叉口操作,来识别关键的可预测性差距,这些行为传统安全过滤器忽视。这项工作为定义统计安全闸门提供了数学严谨的基础,使能够对自动驾驶车队的安全部署进行客观、数据驱动的验证。

英文摘要:

Safety validation for Level 4 autonomous vehicles (AVs) is currently bottlenecked by the inability to scale the detection of rare, high-risk long-tail scenarios using traditional rule-based heuristics. We present Deep-Flow, an unsupervised framework for safety-critical anomaly detection that utilizes Optimal Transport Conditional Flow Matching (OT-CFM) to characterize the continuous probability density of expert human driving behavior. Unlike standard generative approaches that operate in unstable, high-dimensional coordinate spaces, Deep-Flow constrains the generative process to a low-rank spectral manifold via a Principal Component Analysis (PCA) bottleneck. This ensures kinematic smoothness by design and enables the computation of the exact Jacobian trace for numerically stable, deterministic log-likelihood estimation. To resolve multi-modal ambiguity at complex junctions, we utilize an Early Fusion Transformer encoder with lane-aware goal conditioning, featuring a direct skip-connection to the flow head to maintain intent-integrity throughout the network. We introduce a kinematic complexity weighting scheme that prioritizes high-energy maneuvers (quantified via path tortuosity and jerk) during the simulation-free training process. Evaluated on the Waymo Open Motion Dataset (WOMD), our framework achieves an AUC-ROC of 0.766 against a heuristic golden set of safety-critical events. More significantly, our analysis reveals a fundamental distinction between kinematic danger and semantic non-compliance. Deep-Flow identifies a critical predictability gap by surfacing out-of-distribution behaviors, such as lane-boundary violations and non-normative junction maneuvers, that traditional safety filters overlook. This work provides a mathematically rigorous foundation for defining statistical safety gates, enabling objective, data-driven validation for the safe deployment of autonomous fleets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑