检索结果-内蒙古大学图书馆

Dual-Layered Model Protection Scheme Against Backdoor Attacks in Fog Computing-Based Federated Learning

IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT 2025年第2期22卷 2000-2016页

作者： Gu, Ke Zuo, Yiming Tan, Jingjing Yin, Bo Yang, Zheng Li, Xiong Changsha Univ Sci & Technol Sch Comp & Commun Engn Changsha 410114 Peoples R China Univ Elect Sci & Technol China Inst Cyber Secur Sch Comp Sci & Engn Chengdu 611731 Peoples R China

With the growing popularity of federated learning, the security of training models against backdoor attacks has become a key challenge. Existing defense schemes often fail to address the complexity and diversity of such attacks so as to make training models vulnerable. In this paper, we propose a comprehensive dual-layered model protection scheme for fog computing-based federated learning framework. In our scheme, we first introduce a multi-metric defense mechanism deployed on fog servers to defend against malicious backdoor attacks from edge devices. The proposed defense mechanism employs multiple detection indicators to simultaneously evaluate and detect gradient and model training attributes, so that the abnormal local gradients are identified effectively. Further, we construct a second-layered defense scheme deployed on aggregation servers to regularly monitor the participation status of fog servers, whose purpose is to detect the distribution of uploaded gradients and eliminate malicious gradients from compromised fog servers. Additionally, we design an adaptive gradient adjustment method to mitigate the influence of deleting malicious gradients on the global model training process. Experimental results show that our dual-layered model protection scheme can perform well against three type of backdoor attacks (BadNet, Blended and WaNet).

关键词： Servers Computational modeling Training Data models Adaptation models Protection Federated learning Security image edge detection Analytical models backdoor attacks defence dual-layered model protection

来源：评论

学校读者我要写书评

暂无评论

Lightweight Gaussian Process-Based Visual Servoing for Autonomous Wheelchair Sidewalk Navigation

引用

IEEE ACCESS 2025年 13卷 69582-69595页

作者： Hafez, A. H. Abdul Osman, Ismail Haj Ugur, Efgan Kara, Tolgay King Faisal Univ Fac Comp Sci & Informat Technol Dept Comp Sci Al Hasa 31982 Saudi Arabia Univ Tulsa Dept Elect & Comp Engn Tulsa OK 74104 USA Nigde Omer Halisdemir Univ Fac Engn Dept Comp Engn TR-51240 Nigde Turkiye Gaziantep Univ Fac Engn Dept Elect & Elect Engn TR-27310 Gaziantep Turkiye

The personal mobility of wheelchair users is pivotal to their overall well-being. Navigating electric wheelchairs through confined spaces poses notable challenges, particularly for individuals with disabilities. In response, this research introduces a pioneering vision-based approach for sidewalk navigation that leverages tactile paving features and advanced machine learning models. While many existing solutions rely on expensive sensors or GPU-accelerated deep networks, we address this limitation by employing only a low-cost monocular camera and a lightweight GP model. Running at 7 Hz on a Raspberry Pi, our approach offers real-time autonomy without requiring specialized or high-end hardware. The proposed methodology includes creating a specialized dataset, formulating a custom control law, and deploying a lightweight real-time Gaussian Process (GP) model on a Raspberry Pi platform. Thorough experimentation substantiates the efficacy of the presented approach, demonstrating accurate and dependable autonomous wheelchair navigation on sidewalks. Beyond its technical contributions, the proposed solution offers a cost-effective alternative to conventional sensor-dependent systems, profoundly enhancing user mobility and quality of life. By equipping wheelchair users with enhanced navigation capabilities, this research strives to foster greater independence and inclusion in their daily lives.

关键词： Wheelchairs Navigation Cameras Feature extraction Visual servoing Visualization Robots Real-time systems Jacobian matrices image edge detection Gaussian process wheelchair navigation wheelchair control vision-based navigation tactile paving sidewalk navigation

来源：评论

学校读者我要写书评

暂无评论

Acquisition of Phase-Contrast X-Ray images With Commercial Off-the-Shelf CMOS image Sensors

引用

IEEE SENSORS JOURNAL 2025年第6期25卷 9618-9625页

作者： Perez, Martin Lipovetzky, Jose Marpegan, Luciano Haro, Miguel Sofo Blostein, Juan Jeronimo Berisso, Mariano Gomez Crespo, Maria Liz Cicuttin, Andres IMB CSIC CNM Barcelona 08193 Spain Comis Nacl Energia Atom CNEA RA-8400 San Carlos De Bariloche Argentina Consejo Nacl Invest Cient & Tecn C1425 Buenos Aires Argentina Univ Nacl Cuyo Inst Balseiro RA-8400 San Carlos De Bariloche Argentina Abdus Salam Int Ctr Theoret Phys ICTP I-34100 Trieste Italy

This article presents a system to obtain X-ray phase-contrast images using a commercial-off-the-shelf (COTS), low-cost, CMOS image sensor as a direct radiation detector. The 5.2- mu m pixel pitch allows the detection of Fresnel diffraction fringes, which enables the observation of borders and interfaces in samples. A simple experimental setup comprising a microfocus, tungsten anode, X-ray source, an image sensor, and the sample placed at tens of centimeters from the detector is used to measure Fresnel diffraction patterns in opaque and semi-transparent samples on semi-infinite planes. In addition, two biological samples are presented as case studies. The sensor's spectral resolution allows for a better understanding of the results.

关键词： X-ray imaging Sensors Photonics X-ray diffraction Imaging Detectors Blades Diffraction image edge detection Radiography image sensors multispectral imaging photon detection

来源：评论

学校读者我要写书评

暂无评论

引用

IEEE ACCESS 2025年 13卷 19379-19387页

作者： Elsharkawi, Ismael Sharara, Hossam Rafea, Ahmed Amer Univ Cairo Dept Comp Sci & Engn New Cairo 11835 Egypt

image representation in computer vision is a long-standing problem that has a significant impact on any machine learning model performance. There have been multiple attempts to tackle this problem that were introduced in the literature, starting from traditional Convolutional Neural Networks (CNNs) to Vision Transformers and MLP-Mixers that were more recently introduced to represent images as sequences. Most recently, Vision Graph Neural Networks (ViG) have shown very promising performance through representing images as graphs. The performance of ViG models heavily depends on how the graph is constructed. The ViG model relies on k-nearest neighbors (k-nn) for graph construction, which while achieving very good performance on classical computer vision tasks, imposes a number of challenges, such as determining the optimal value for k, as well as using the same chosen value for all nodes in a graph, which in turns reduces the graph expressiveness and limits the power of the model. In this paper, we propose a new approach that relies on similarity score thresholding to create the graph edges and, subsequently, pick the neighboring nodes. Rather than the number of neighbors, we allow for the specification of the normalized similarity threshold as an input parameter for each layer, which is more intuitive. We also propose a decreasing threshold framework to select the input threshold for all layers. We show that our proposed method can achieve higher performance than the ViG model for image classification on the benchmark imageNet-1K dataset, without increasing the complexity of the model.

关键词： Computer vision Computer architecture Vectors Graph neural networks Transformers image edge detection Point cloud compression image resolution image representation image classification Graph Neural Networks Vision Graph Neural Networks image Classification

来源：评论

学校读者我要写书评

暂无评论

Emergence Model of Perception With Global-Contour Precedence Based on Gestalt Theory and Primary Visual Cortex

引用

IEEE TRANSACTIONS ON image PROCESSING 2025年 34卷 2721-2736页

作者： Li, Jingmeng Wei, Hui Fudan Univ Sch Comp Sci Lab Algorithms Cognit Models Shanghai 200438 Peoples R China Fudan Univ Innovat Ctr Callig & Painting Creat Technol Sch Comp Sci Lab Algorithms Cognit ModelsMCT Shanghai 200438 Peoples R China

Perceptual edge grouping is a technique for organizing the cluttered edge pixels into meaningful structures and further serves high-level vision tasks, which has long been a basic and critical task in computer vision. Existing methods usually have a poor performance when coping with the junctions caused by occlusion and noise in natural images. In this paper, we present GPGrouper, a perceptual edge grouping model based on gestalt theory and the primary visual cortex (V1). Different from the existing methods, GPGrouper leverages the edge representation and grouping matrix (ERGM), a functional structure inspired by V1 mechanisms, to represent edges in a way that can effectively reduce grouping errors caused by occlusion between objects. ERGM is trained with natural image contours and further provides a priori guidance for the construction of the edge connection graph (ECG) that is useful to minimize the impact of noise on grouping. In the experiment, we compared GPGrouper and the state-of-the-art (SOTA) method of perceptual grouping on the visual psychology pathfinder challenge. The results demonstrate that GPGrouper outperforms the SOTA method in grouping performance. Furthermore, in the grouping experiments involving line segments with varying lengths detected by the Line Segment Detector (LSD), as well as those involving superpixel segmentation results with significant levels of interfering noise using the SLIC algorithm, GPGrouper was superior to the existing methods in terms of grouping effect and robustness. Moreover, the results of applying the grouping results to the vision tasks objectness demonstrate that GPGrouper can contribute significantly to high-level visual tasks.

关键词： image edge detection Visualization Computational modeling Noise image segmentation Visual systems Psychology Organizations Vectors Object detection Perceptual organization global precedence effect Gestalt theory primary visual cortex perceptual edge grouping

来源：评论

学校读者我要写书评

暂无评论

Decentralized Low-Latency Collaborative Inference via Ensembles on the edge

引用

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS 2025年第1期24卷 598-614页

作者： Malka, May Farhan, Erez Morgenstern, Hai Shlezinger, Nir Ben Gurion Univ Negev Sch ECE IL-8410501 Beer Sheva Israel Facebook Tel Aviv Israel Ramon Space IL-4524075 Hod Hasharon Israel

The success of deep neural networks (DNNs) is heavily dependent on computational resources. While DNNs are often employed on cloud servers, there is a growing need to operate DNNs on edge devices. edge devices are typically limited in their computational resources, yet, often multiple edge devices are deployed in the same environment and can reliably communicate with each other. In this work we propose to facilitate the application of DNNs on the edge by allowing multiple users to collaborate during inference to improve their accuracy. Our mechanism, coined edge ensembles, is based on having diverse predictors at each device, which form an ensemble of models during inference. To mitigate the communication overhead, the users share quantized features, and we propose a method for aggregating multiple decisions into a single inference rule. We analyze the latency induced by edge ensembles, showing that its performance improvement comes at the cost of a minor additional delay under common assumptions on the communication network. Our experiments demonstrate that collaborative inference via edge ensembles equipped with compact DNNs substantially improves the accuracy over having each user infer locally, and can outperform using a single centralized DNN larger than all the networks in the ensemble together.

关键词： Collaboration Servers Accuracy image edge detection Artificial neural networks Reliability Low latency communication Protocols Computational modeling Wireless communication Collaborative inference ensemble learning

来源：评论

学校读者我要写书评

暂无评论

Neural Network Pruning for Lightweight Metal Corrosion image Segmentation Models

引用

IEEE ACCESS 2025年 13卷 71673-71687页

作者： Yu, Vincent F. Santiyuda, Gemilang Lin, Shih-Wei Pasaribu, Udjianna S. Afrianti, Yuli Sri Natl Taiwan Univ Sci & Technol Dept Ind Management Taipei 106335 Taiwan Inst Bisnis & Teknol Indonesia Fac Technol & Informat Denpasar 80225 Bali Indonesia Chang Gung Univ Dept Informat Management Taoyuan 333 Taiwan Ming Chi Univ Technol Dept Ind Engn & Management Taipei 243 Taiwan Keelung Chang Gung Mem Hosp Dept Emergency Med Keelung 203 Taiwan Bandung Inst Technol Fac Math & Nat Sci Stat Res Grp Bandung 40132 Indonesia

Metal corrosion detection is essential for ensuring structural safety and minimizing economic losses. While deep learning (DL)-based image segmentation has improved corrosion detection accuracy and efficiency, its high computational demands hinder deployment on resource-constrained edge devices. This study investigates lightweight DL models for corrosion segmentation by applying structured pruning to reduce computational costs while maintaining accuracy. We evaluate five segmentation architectures (U-Net, U-Net++, FPN, LinkNet, and MA-Net) and three pruning strategies (linear, automated gradual pruning, and movement pruning) on two corrosion image datasets (NEA and SSCS). Detailed trade-off analysis between model size, computational cost (MAC), and segmentation performance (IoU) reveals that pruning up to 90% sparsity leads to a <= 10% IoU drop on SSCS and <= 5% on NEA, demonstrating that significant compression is possible with minimal accuracy loss. However, some architectures (e.g., LinkNet) and pruning strategies (e.g., movement pruning) show significant performance deterioration, suggesting that pruning effectiveness varies across models. These findings provide insights into optimizing corrosion segmentation models for efficient deployment on edge devices, balancing accuracy and resource constraints.

关键词： Corrosion image segmentation Metals Monitoring Computational modeling Computer architecture Accuracy Optimization image edge detection Biological system modeling Computer vision deep learning image segmentation metal corrosion pruning

来源：评论

学校读者我要写书评

暂无评论

DGFEG: Dynamic Gate Fusion and edge Graph Perception Network for Remote Sensing Change detection

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2025年 18卷 3581-3598页

作者： Zhou, Shengning Yuan, Genji Hua, Zhen Li, Jinjiang Shandong Technol & Business Univ Sch Informat & Elect Engn Yantai 264005 Peoples R China Shandong Technol & Business Univ Sch Comp Sci & Technol Yantai 264005 Peoples R China

Benefiting from continuous innovations in deep learning algorithms, the accuracy of building change detection (BCD) in remote sensing (RS) has significantly improved. Numerous networks combining CNN and transformer architectures have emerged, yet effectively balancing local detail and global context features remains a topic of ongoing discussion. Furthermore, accurately leveraging edge information within RS images to enhance the recognition of structural changes in buildings is another critical challenge. To address these issues, this article proposes a BCD network based on dynamic gate fusion and edge graph perception (DGFEG). First, a hybrid backbone, MCTrans, is employed as the encoder to extract multiscale detailed features and global positional information of buildings. Second, a dynamic gate fusion module is introduced to dynamically weight and fuse the concatenated and differential features obtained by the encoder, enhancing the semantic representation of actual building change regions. Finally, an edge graph perception module integrates edge information with the fused features, leveraging the spatial similarity of graph structures and the interaction of edge features to suppress irrelevant edge interference, thereby improving the model's sensitivity and accuracy in detecting subtle building changes. In experiments, DGFEG was tested on real-world change scenarios and multiple RSCD datasets. The results demonstrate its superior performance compared to existing state-of-the-art methods, proving its excellence and broad application potential in tackling complex BCD tasks.

关键词： Buildings Feature extraction image edge detection Accuracy Remote sensing Logic gates Convolutional neural networks Sensitivity Semantics Noise Attention mechanism building change detection (BCD) dynamic gate fusion (DGF) edge graph perception (EGP) remote sensing (RS) images

来源：评论

学校读者我要写书评

暂无评论

Enhancing Single-Cell RNA-Seq Data Completeness With a Graph Learning Framework

引用

IEEE TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS 2025年第1期22卷 64-72页

作者： Lall, Snehalika Ray, Sumanta Bandyopadhyay, Sanghamitra Indian Stat Inst Machine Intelligence Unit Kolkata 700108 West Bengal India Aliah Univ Kolkata 700156 West Bengal India

Single cell RNA sequencing (scRNA-seq) is a powerful tool to capture gene expression snapshots in individual cells. However, a low amount of RNA in the individual cells results in dropout events, which introduce huge zero counts in the single cell expression matrix. We have developed VAImpute, a variational graph autoencoder based imputation technique that learns the inherent distribution of a large network/graph constructed from the scRNA-seq data leveraging copula correlation ($Ccor$) among cells/genes. The trained model is utilized to predict the dropouts events by computing the probability of all non-edges (cell-gene) in the network. We devise an algorithm to impute the missing expression values of the detected dropouts. The performance of the proposed model is assessed on both simulated and real scRNA-seq datasets, comparing it to established single-cell imputation methods. VAImpute yields significant improvements to detect dropouts, thereby achieving superior performance in cell clustering, detecting rare cells, and differential expression.

关键词： Imputation Gene expression Correlation Protocols Predictive models Mathematical models RNA image edge detection Bioinformatics Accuracy Single cell imputation variational graph autoencoder copula RgCop

来源：评论

学校读者我要写书评

暂无评论

A Novel Intensity-Corrected Blue Channel Compensation and edge-Preserving Contrast Enhancement Using Laplace Filter and Sigmoid Function for Sand-Dust image Enhancement

引用

IEEE ACCESS 2025年 13卷 43127-43144页

作者： Masood, Muhammad Khawaja Kashif Baro, Enrique Nava Roth, Pablo Otero Univ Malaga Inst Ocean Engn Res Malaga 29071 Spain Qassim Univ Elect Engn Dept Buraydah 52571 Saudi Arabia

Outdoor computer vision systems face significant challenges due to reduced visibility and severe color distortion in the images captured in sand-dust-affected environments. This study aims to improve the visibility of sand-dust-degraded images. To achieve this goal, a novel and effective method is proposed to remove the sand-dust color cast and enhance image visibility. The proposed method combines two essential color model methods to remove the sand-dust color cast and enhance image clarity. In the initial phase, sand-dust removal is achieved using a novel Intensity-corrected blue channel compensation along with white balancing for color adjustment based on the Red-Green-Blue (RGB) color model. In the next phase, a novel edge-preserving contrast enhancement method is applied to improve the visibility under sand-dust conditions. This method consists of CLAHE, a Gaussian blur filter, a Laplace filter, and the sigmoid function. Using the Hue-Saturation-Value (HSV) color model, CLAHE is applied for contrast enhancement;the Gaussian blur filter removes high-frequency noise, and the Laplace filter enhances edge detection, all targeting the V (Value) channel to refine image details, while the sigmoid function adjusts saturation in the Saturation (S) channel, ensuring natural color balance and improved feature visibility. In-depth qualitative and quantitative evaluations are conducted on images with varying levels of sand-dust intensity (weak, moderate, strong, extreme). The proposed method shows superior performance in Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and processing speed, while significantly reducing computational complexity. Compared to the state-of-the-art CNN and all previous methods, our proposed method is efficient for real-time applications with minimal hardware requirements, making it ideal for embedded vision systems. Furthermore, a novel Energy Efficiency Index (EEI) is used to assess computational cost-effectiveness. The ev

关键词： image color analysis image edge detection image restoration Visualization Distortion Computational efficiency Green products Lighting Indexes Energy efficiency Sand-dust images intensity corrected blue channel compensation contrast enhancement Laplace filter sigmoid function CLAHE Gaussian blur filter

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：