用语言描述局部线索，提升边界附近异常检测精度
PWLR: Pairwise Witness Local Rejection for Boundary-Aware Out-of-Distribution Detection
- 通过多模态模型提取支持某类而非另一类的局部语言线索
- 在真实数据上验证，对近域与远域异常检测均有效提升
- 适合需要高精度边界判断的视觉分类场景
图像分类器在处理分布外（OOD）样本时仍面临挑战，尤其当近域样本靠近分布内（ID）类别边界时。现有视觉-语言检测方法通过类别语义、局部提示或大模型生成的异常概念改进检测，但很少利用语言作为区分混淆类别的明确边界证据。本文提出成对见证局部拒绝（PWLR），使用预训练多模态大模型（MLLM）离线描述支持某一类而非特定竞争类的可见局部线索，并在仅含分布内数据的条件下，通过冻结的视觉-语言主干筛选出可靠的局部验证器。推理时，先保留一组全局合理的类别，再检查其中是否存在对最相关竞争类的局部支持，最后通过校准将成对局部证据与全局类别得分结合。在ImageNet-100远域、清洁/挑战性域及近域基准上的实验表明，PWLR在多个主干网络上持续提升强基线性能。源代码将公开。
原文摘要 · Abstract (English)
Out-of-distribution (OOD) detection remains challenging for image classifiers, especially when near-OOD samples lie close to in-distribution (ID) class boundaries. Recent vision-language detectors improve OOD detection through class semantics, local prompting, or LLM-generated outlier concepts, but seldom use language as explicit boundary evidence between confusing ID classes. We propose Pairwise Witness Local Rejection (PWLR), which uses an MLLM offline to describe visible local cues that favor one ID class over a specific rival class. These cue phrases are then screened with ID-only data under a frozen vision-language backbone, so that only reliable local verifiers are kept. At inference, PWLR first retains a small set of globally plausible classes, then checks whether any of them is locally supported against its most relevant rivals, and finally combines this pairwise local evidence with the global class score through calibration. Experiments on ImageNet-100 far-OOD, cleaner/challenging OOD and near-OOD benchmarks show that PWLR consistently improves strong vision-language baselines across multiple backbones. Source code will be released.
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。