构建大规模持续异常检测基准，提升模型对未知异常的泛化能力。
Continual-MEGA: A Large-scale Benchmark for Generalizable Continual Anomaly Detection
- 融合多个数据集构建新基准，支持持续学习与零样本泛化。
- 现有方法在像素级缺陷定位上仍有明显不足。
- 适合关注工业质检、自适应异常检测的研究者。
本文提出一个面向持续异常检测的大规模基准框架Continual-MEGA，旨在更贴近真实应用场景。该基准通过整合精心筛选的现有数据集与新提出的ContinualAD数据集，显著扩展了评估场景。除常规的连续学习增量设置外，还引入了衡量未见类别零样本泛化能力的新任务，要求模型在持续适应中同时提升对未知异常的识别性能。我们还提出统一基线算法，在少样本检测中表现更稳健，并保持强泛化性。大量实验揭示三个关键发现：(1) 现有方法在像素级缺陷定位方面仍有巨大改进空间；(2) 所提方法始终优于已有方法；(3) 新增的ContinualAD数据集能有效提升主流异常检测模型的表现。相关代码与数据已开源至https://github.com/Continual-Mega/Continual-Mega。
原文摘要 · Abstract (English)
In this paper, we introduce a new benchmark for continual learning in anomaly detection, aimed at better reflecting real-world deployment scenarios. Our benchmark, Continual-MEGA, includes a large and diverse dataset that significantly expands existing evaluation settings by combining carefully curated existing datasets with our newly proposed dataset, ContinualAD. In addition to standard continual learning with expanded quantity, we propose a novel scenario that measures zero-shot generalization to unseen classes, those not observed during continual adaptation. This setting poses a new problem setting that continual adaptation also enhances zero-shot performance. We also present a unified baseline algorithm that improves robustness in few-shot detection and maintains strong generalization. Through extensive evaluations, we report three key findings: (1) existing methods show substantial room for improvement, particularly in pixel-level defect localization; (2) our proposed method consistently outperforms prior approaches; and (3) the newly introduced ContinualAD dataset enhances the performance of strong anomaly detection models. We release the benchmark and code in https://github.com/Continual-Mega/Continual-Mega.
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。