arXiv 2401.010102024-01-02cs.CVcs.LG

基于对比学习提示词的无监督持续异常检测

Unsupervised Continual Anomaly Detection with Contrastively-learned Prompt

Southern University of Science and Technology（南方科技大学）；Tencent Youtu Lab（腾讯优图实验室）；Shanghai Jiao Tong University（上海交通大学）

Jiaqi Liu, Kai Wu, Qiang Nie, Ying Chen, Bin-Bin Gao, Yong Liu, Jinbao Wang, Chengjie Wang, Feng Zheng

备注：: Accepted by AAAI 2024

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对工业制造中无监督异常检测的灾难性遗忘问题，提出UCAD框架，通过对比学习提示词和结合SAM的基于结构对比学习，赋予模型持续学习能力，显著优于现有异常检测方法。

摘要：

具有增量训练的无监督异常检测（UAD）在工业制造中至关重要，因为不可预测的缺陷使得获取充足的标注数据变得不可行。然而，持续学习方法主要依赖有监督标注，而由于缺乏监督，其在UAD中的应用受到限制。当前的UAD方法按顺序为不同类别训练单独的模型，导致灾难性遗忘和沉重的计算负担。为解决这一问题，我们提出了一种名为UCAD的新型无监督持续异常检测框架，该框架通过对比学习的提示词赋予UAD持续学习能力。在提出的UCAD中，我们设计了一个持续提示模块（CPM），利用简洁的键-提示-知识记忆库，通过特定任务的“正常”知识来引导任务不变的“异常”模型预测。此外，设计了基于结构的对比学习（SCL），结合Segment Anything Model（SAM）来改进提示学习和异常分割结果。具体而言，通过将SAM的掩码作为结构，我们将同一掩码内的特征拉近，并将其他特征推开，以获得通用的特征表示。我们进行了全面实验，并设定了无监督持续异常检测和分割的基准，证明我们的方法显著优于异常检测方法，甚至优于使用重放训练的方法。代码将发布于https://github.com/shirowalker/UCAD。

英文摘要：

Unsupervised Anomaly Detection (UAD) with incremental training is crucial in industrial manufacturing, as unpredictable defects make obtaining sufficient labeled data infeasible. However, continual learning methods primarily rely on supervised annotations, while the application in UAD is limited due to the absence of supervision. Current UAD methods train separate models for different classes sequentially, leading to catastrophic forgetting and a heavy computational burden. To address this issue, we introduce a novel Unsupervised Continual Anomaly Detection framework called UCAD, which equips the UAD with continual learning capability through contrastively-learned prompts. In the proposed UCAD, we design a Continual Prompting Module (CPM) by utilizing a concise key-prompt-knowledge memory bank to guide task-invariant `anomaly' model predictions using task-specific `normal' knowledge. Moreover, Structure-based Contrastive Learning (SCL) is designed with the Segment Anything Model (SAM) to improve prompt learning and anomaly segmentation results. Specifically, by treating SAM's masks as structure, we draw features within the same mask closer and push others apart for general feature representations. We conduct comprehensive experiments and set the benchmark on unsupervised continual anomaly detection and segmentation, demonstrating that our method is significantly better than anomaly detection methods, even with rehearsal training. The code will be available at https://github.com/shirowalker/UCAD.

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。