用空间逻辑推理提升道路异常检测，减少误报且无需多模型堆叠。
LARAD: Layout-Aware Road Anomaly Detection via Spatial-Logic Reasoning
- 通过生成空间不合理但纹理一致的样本，训练模型识别上下文违例。
- 在真实道路数据集上达到新最优性能，误报率显著降低。
- 轻量级设计适合实时自动驾驶系统部署。
精准的开放世界障碍物检测对自动驾驶至关重要。现有异常分割方法存在根本缺陷：过度依赖纹理新颖性识别分布外（OoD）物体，忽视上下文空间逻辑。此外，消除误报常需级联大量视觉模型，导致不可接受的推理延迟。为此，我们提出布局感知道路异常检测（LARAD），将范式从外观匹配转向空间逻辑推理。首先，引入空间逻辑违规合成（SLVS）流程，生成纹理一致但空间无效的训练样本，迫使模型学习上下文违例。其次，为标准封闭集分割网络增加一个轻量级、面向分布外的注意力分支。大量实验表明，LARAD显著提升对逻辑异常的鲁棒性，并建立新基准，同时保持单模型架构的高效率。
原文摘要 · Abstract (English)
Accurate open-world obstacle detection is critical for autonomous driving. Current anomaly segmentation methods suffer from a fundamental blind spot: they over-rely on texture novelty to identify out-of-distribution (OoD) objects while ignoring contextual spatial logic. Furthermore, mitigating the resulting false positives often requires cascading massive vision models, introducing unacceptable inference latency. To address these issues, we propose Layout-Aware Road Anomaly Detection (LARAD), shifting the paradigm from appearance matching to spatial-logic reasoning. First, we introduce the Spatial-Logic Violation Synthesis (SLVS) pipeline, which generates training samples that are texture-consistent yet spatially invalid, forcing the model to learn contextual violations. Second, we augment a standard closed-set segmentation network with a lightweight, OoD-guided attention branch. Extensive experiments demonstrate that LARAD significantly enhances robustness against logical anomalies and establishes a new state-of-the-art, all while retaining the high efficiency of a single-model architecture.
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。