检索结果-内蒙古大学图书馆

MemFlow-AD: An anomaly detection and localization model based on memory module and normalizing flow

JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION 2025年 110卷

作者： Xie, Xin Li, Zixi Xiong, Shenping Liu, Zhaoyang Cai, Tijian East China Jiaotong Univ Sch Informat & Software Engn Nanchang 330013 Jiangxi Peoples R China

Presently, in most anomaly detection methods, the training dataset contains a low frequency of anomalous data with diverse categories. However, these methods exhibit limited learning capacity for anomalous information, leading to weak model generalization and low detection accuracy. This paper proposes MemFlow-AD, an anomaly detection and localization model that integrates a memory module and a normalizing flow. MemFlowAD supplements anomalous data using anomaly simulation, retains general patterns from normal samples via the memory module to discern potential differences between normal and anomalous samples, employs a 2D normalizing flow to extract distributional feature information from the data, and through multiscale feature fusion and attention mechanism to further enhance the feature expression ability of the model. Experimental results demonstrate outstanding performance in detecting and localizing anomalies on the MVTec dataset, achieving accuracies of 98.61% and 94.02%, respectively. Moreover, on the BTAD dataset, the model exhibits a 2.15% improvement in detection accuracy compared to current mainstream methods.

关键词： anomaly detection and localization Memory module Normalizing flow Attention mechanism

来源：评论

学校读者我要写书评

暂无评论

Multiscale Prototype Fusion Network for Industrial Product Surface anomaly detection and localization

引用

IEEE SENSORS JOURNAL 2024年第20期24卷 32707-32716页

作者： Shao, Haidong Peng, Jiangji Shao, Minghui Liu, Bin Hunan Univ Coll Mech & Vehicle Engn Changsha 410082 Peoples R China Univ Strathclyde Dept Management Sci Glasgow G1 1XG Scotland

In complex industrial application scenarios, abnormal samples are scarce. In the case of weak defect features, the high similarity of positive and negative samples further complicates detection and localization. In addition, anomalies are often subtle and unpredictable, which makes it particularly difficult to detect and localize anomalous subregions with unknown anomaly patterns. Many detection algorithms suffer from high computational complexity and huge memory consumption. To address these challenges, this article proposes a multiscale prototype fusion network (MPFnet) for industrial product surface anomaly detection and localization. MPFnet uses multiscale prototypes to construct representative normal patterns and incorporates a multiscale fusion block to facilitate information exchange between different scales. This design enhances the model's attention to characterizing the prototype and normal features. Feature adapter is constructed to generate fitness features, reducing domain bias. By adding noise to the adapted features, anomalous features are generated, and anomalies are detected using a simple and efficient discriminator. A large number of experiments were carried out on the challenging MVTec AD and MVTec LOCO AD datasets, demonstrating that MPFnet outperforms other state-of-the-art comparative methods, achieving good detection and localization results regardless of defect patterns.

关键词： Feature extraction Image reconstruction Prototypes Location awareness anomaly detection Training Adaptation models anomaly detection and localization feature adapter industrial application scenarios multiscale information exchange prototype fusion network

来源：评论

学校读者我要写书评

暂无评论

Residual Attention Dual Autoencoder for anomaly detection and localization in Cigarette Packaging

Residual Attention Dual Autoencoder for Anomaly Detection an...

引用

Chinese Automation Congress (CAC)

作者： Zhu, Liming Zhang, Qiang Wang, Wei Hangzhou Cigarette Factory China Tobacco Zhejiang Hangzhou Peoples R China

ISBN: (纸本)9781728176871

anomaly detection and localization is an important problem in the industrial production process, such as cigarette packaging quality inspection. This task is challenging due to very few abnormal samples. This paper proposes Residual Attention Dual Autoencoder (RADAE) composed of the dual autoencoder and the attention map generator. Without the need of the abnormal training images, the proposed model can achieve anomaly detection and localization by unsupervised learning. The dual autoencoder is an encoder-decoder-encoder network that maps the input image to a lower dimension vector for reconstruction, and remaps the generated image to its latent representation. The data distribution of the normal samples is learned by training the dual autoencoder. anomaly discrimination is realized by measuring the difference between the latent space of the original image and the generated image during detection. And the attention map generator is a CNN model, which uses the residuals of the original image and the generated image to obtain the attention map under the guidance of the attention loss function. The generated attention map is encouraged to focus on all regions in normal samples during training, and spontaneously focus on abnormal regions to achieve localization during testing. The experimental results demonstrate the reliability and superior performance of the proposed model for anomaly detection and localization in the cigarette packaging dataset.

关键词： anomaly detection and localization cigarette packaging unsupervised learning autoencoder attention map

来源：评论

学校读者我要写书评

暂无评论

SOM-Based anomaly detection and localization for Space Subsystems 13th

SOM-Based Anomaly Detection and Localization for Space Subsy...

引用

13th International Workshop on Self-Organizing Maps, Learning Vector Quantization, Clustering and Data Visualization (WSOM)

作者： Rosengarten, Maia Ramachandran, Sowmya Stottler Henke Associates Inc San Mateo CA 94402 USA

ISBN: (纸本)9783030196417;9783030196424

The aim of this paper is to contribute to machine-learning technology that expands real-time and offline Integrated System Health Management capabilities for future deep-space exploration efforts. To this end, we have developed anomaly detection via Topological feature-Map (ADTM), which leverages a Self-Organizing Map (SOM)-based architecture to produce high-resolution clusters of nominal system behavior. What distinguishes ADTM from more common clustering techniques (e.g. k-means) is that it maps high-dimensional input vectors to a 2D grid while preserving the topology of the original dataset. The result is a 'semantic map' that serves as a powerful tool for uncovering latent relationships between features of the incoming data points. We successfully modeled and analyzed datasets from a NASA Ames Research Center Graywater Recycling System which documents a real hardware system fault. Our results show that ADTM effectively detects both known and unknown anomalies and identifies the correlated measurands from models trained using just nominal data.

关键词： Self-Organizing Map anomaly detection and localization Integrated System Health Management

来源：评论

学校读者我要写书评

暂无评论

Unsupervised anomaly detection and localization Based on Two-Hierarchy Normalizing Flow

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2024年 73卷 1页

作者： Jiang, Jielin Wei, Shun Xu, Xiaolong Cui, Yan Liu, Xiying Nanjing Univ Informat Sci & Technol Sch Software Jiangsu Prov Engn Res Ctr Adv Comp & Intelligent S Nanjing 210044 Peoples R China Nanjing Univ Informat Sci & Technol Jiangsu Collaborat Innovat Ctr Atmospher Environm Nanjing 210044 Peoples R China Nanjing Univ State Key Lab Novel Software Technol Nanjing 210023 Peoples R China Nanjing Univ Informat Sci & Technol Sch Software Nanjing 210044 Peoples R China Nanjing Normal Univ Special Educ Coll Math & Informat Sci Nanjing 210038 Peoples R China

Unsupervised anomaly detection (UAD) methods are widely used in industrial anomaly detection, primarily since there is a lack of anomalous data available for training. However, these methods still struggle to effectively detect and localize anomalies due to the diverse types of anomalies and frequent variations in their sizes and shapes. To address this challenge, we propose a new flow-based framework called two-hierarchy normalizing flow (THF) for anomaly detection and localization. THF consists of two main components: a masked flow and a constant flow. In the masked flow, we improve the traditional convolutional kernel by masking partial regions of the kernel to prevent excessive minimization of negative log-likelihood. This enhancement greatly helps in effectively detecting and localizing anomalies. The constant flow incorporates dual units (DUs) and a volume-preserving flow (VPF) module. The DUs consist of a characterization unit and a mixture unit. The characterization unit accurately captures both local and global feature information, while the mixture unit learns neighboring feature representations. Unlike traditional flow-based methods, the VPF module maintains the volume of the probability distribution invariable, facilitating precise detection and localization of high-scoring anomalies. Extensive experiments conducted on multiple widely used anomaly detection datasets demonstrate that THF outperforms state-of-the-art (SOTA) methods in both anomaly detection and localization.

关键词： anomaly detection Feature extraction Location awareness Noise measurement Image reconstruction Computational modeling Training anomaly detection and localization deep learning mask normalizing flow (NF) unsupervised

来源：评论

学校读者我要写书评

暂无评论

anomaly REGION detection AND localization IN METAL SURFACE INSPECTION 23

ANOMALY REGION DETECTION AND LOCALIZATION IN METAL SURFACE I...

引用

23rd IEEE International Conference on Image Processing (ICIP)

作者： Vaikundam, Sriram Hung, Tzu-Yi Chia, Liang Tien Nanyang Technol Univ Corp Lab Rolls Royce Singapore Singapore Nanyang Technol Univ Sch Comp Engn Singapore Singapore

ISBN: (纸本)9781467399616

Visual inspection and identification of anomalies are important steps in the manufacturing process. A data set containing images of metallic components at different orientations and lighting conditions have been used for this experiment. Our aim is to find the anomaly in an image by subtracting it from an equivalent reference image with no anomalies. This is challenging due to the availability of multiple and slightly changing viewpoints for the reference image. The anomalies can only be detected in the residue obtained by using an identical reference image with no anomalies in the subtraction process. The proposed system is based on finding the best reference from a pool of similar images. A method for ranking the images based on similarity is also introduced. The proposed method scales well for high contrast images with dark colored anomalies.

关键词： Metal surface inspection anomaly detection and localization

来源：评论

学校读者我要写书评

暂无评论

anomaly REGION detection AND localization IN METAL SURFACE INSPECTION

ANOMALY REGION DETECTION AND LOCALIZATION IN METAL SURFACE I...

引用

IEEE International Conference on Image Processing

作者： Sriram Vaikundam Tzu-Yi Hung Liang Tien Chia Rolls-Royce @ Nanyang Technological University Corporate Lab School of Computer Engineering Nanyang Technological University

ISBN: (纸本)9781467399623

关键词： Metal surface inspection anomaly detection and localization Anomalies images anomaly IMAGE CONTRAST METAL SURFACE Imagery (Psychotherapy) Subtraction Visual inspection Image POOLS

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Framework for Measurement Causality Extraction and FDIA localization

引用

IEEE TRANSACTIONS ON SMART GRID 2025年第3期16卷 2587-2598页

作者： Wu, Shengyang Yang, Chen Wang, Jingyu Shi, Dongyuan Huazhong Univ Sci & Technol State Key Lab Adv Electromagnet Technol Wuhan 430074 Hubei Peoples R China

False Data Injection Attack (FDIA) has become a growing concern in modern cyber-physical power systems. Many learning-based approaches have utilized the statistical correlation patterns between measurements to facilitate the detection and localization of FDIA. However, these correlation patterns are susceptible to the distribution drift of measurement data, which can be induced by changes in system operating points or variations in attack strength, leading to degraded model performance. Causal inference serves as a promising solution to this problem, as it can embed the physical relationship between measurements as causal patterns that are robust against data distribution drifts. However, causal inference is also computationally demanding. To leverage its advantages and address the computational cost issue, this paper proposes a lightweight framework based on causal inference and Graph Attention Networks (GATs) to extract causal patterns between measurements and locate FDIAs. The proposed framework consists of two levels. The lower level uses an X-learner algorithm to estimate the causality strength between measurements and generate Measurement Causality Graphs (MCGs). The upper level then applies a GAT to identify the anomaly patterns in the MCGs. Since the extracted causality patterns are intrinsically related to the measurements, it is easier for the upper level model to identify the attacked nodes than the existing FDIA localization approaches. A physical neighbor masking strategy is implemented to cut down the computational cost of both levels. The performance of the proposed framework is evaluated on the IEEE 39-bus and 118-bus systems. Experimental results show that the causality-based FDIA localization mechanism provides a lightweight solution to interpretable measurement causality extraction and robust FDIA localization.

关键词： Location awareness Cause effect analysis Power systems Correlation Power measurement Computational modeling Training Computational efficiency Vectors Training data Causal inference false data injection attack anomaly detection and localization graph attention network

来源：评论

学校读者我要写书评

暂无评论

SAC-BL: A hypothesis testing framework for unsupervised visual anomaly detection and location

引用

NEURAL NETWORKS 2025年 185卷 107147页

作者： Ma, Xinsong Wu, Jie Liu, Weiwei Wuhan Univ Sch Comp Sci 299 Ba Yi Rd Wuhan 430072 Hubei Peoples R China

Reconstruction-based methods achieve promising performance for visual anomaly detection (AD), relying on the underlying assumption that the anomalies cannot be accurately reconstructed. However, this assumption does not always hold, especially when suffering weak anomalous (a.k.a. normal-like) examples. More significantly, the existing methods primarily devote to obtaining the strong discriminative score functions, but neglecting the systematic investigation of the decision rule based on the proposed score function. Unlike previous work, this paper solves the AD issue starting from the decision rule within the statistical framework, providing a new insight for AD community. Specifically, we frame the AD task as a multiple hypothesis testing problem, Then, we propose a novel betting-like (BL) procedure with an embedding of strong anomaly constraint network (SACNet), called SAC-BL, to address this testing problem. In SAC-BL, BL procedure serves as the decision rule and SACNet is trained to capture the critical discriminative information from weak anomalies. Theoretically, our SAC-BL can control false discovery rate (FDR) at the prescribed level. Finally, we conduct extensive experiments to verify the superiority of SAC-BL over previous method.

关键词： anomaly detection and localization Reconstruction Weak anomaly Hypothesis testing

来源：评论

学校读者我要写书评

暂无评论

Cloud Telemetry Modeling via Residual Gauss-Markov Random Fields 24

Cloud Telemetry Modeling via Residual Gauss-Markov Random Fi...

引用

24th Conference on Innovation in Clouds, Internet and Networks (ICIN)

作者： Landolfi, Nicholas C. O'Neill, Daniel C. Lall, Sanjay Stanford Univ Dept Comp Sci Stanford CA 94305 USA Stanford Univ Dept Elect Engn Stanford CA 94305 USA

ISBN: (纸本)9781728177052

Can probabilistic graphical models characterize cloud telemetry? This paper promotes the affirmative view. Cloud systems are large, connected, and dynamic. Consequently, data-based techniques to model their telemetry are high-dimensional, spatial, and unsupervised. Undirected probabilistic graphical models seem natural, but remain unexplored. We discuss one way around the limitation that cloud measurements violate usual assumptions of normality, and give a tractable estimation procedure for a candidate data model. As a preliminary test, we fit the model and use it to detect and localize anomalies in a synthetic environment and for a small-scale software system.

关键词： cloud performance monitoring anomaly detection and localization probabilistic graphical models Gauss-Markov random fields Markov networks

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：