检索结果-内蒙古大学图书馆

A Signature Recognition Technique With a Powerful Verification Mechanism Based on CNN and PCA

IEEE ACCESS 2024年 12卷 40634-40656页

作者： Abosamra, Gibrael Oqaibi, Hadi King Abdulaziz Univ Fac Comp & Informat Technol Dept Comp Informat Syst Jeddah 21589 Saudi Arabia

In this paper, we embed a signature verification mechanism in a previously introduced architecture for signature recognition to detect in-distribution and out-of-distribution random forgeries. In the previous architecture, a CNN was trained on the genuine user training dataset and then used as a feature extraction module. A k-NN algorithm with cosine distance was then used to classify the unknown signatures based on the nearest cosine distance neighbor. This architecture led to higher than 99% accuracy, but without verification, because any unknown signature will converge to one of the identities of the training dataset's users. To add a verification mechanism that differentiates between genuine and random forgeries, we use PCA to select the most discriminating features used in calculating the cosine distance between the training and testing signatures. A fixed parameter thresholding technique based on the training distances is introduced that best differentiates between the genuine and random-user signatures. Moreover, enhancement of the technique is carried out by combining the output of the Softmax layer and the last convolution layer of the ResNet18 model to get a highly discriminative representation of the handwritten signatures. Accordingly, the introduced verification mechanism resulted in very low false positive and negative rates for test signatures from inside and outside the main dataset, with an insignificant decrease in the high identification accuracy. The complete architecture has been tested on three publicly available datasets, showing superior results.

关键词： Training Principal component analysis Feature extraction Forgery Vectors Convolutional neural networks Data models Nearest neighbor methods detection algorithms Handwriting recognition cosine distance false negative rate false positive rate k-Nearest neighbor algorithm out-of-distribution detection signature identification signature verification

来源：评论

学校读者我要写书评

暂无评论

HGNN-QSSA: Heterogeneous Graph Neural Networks With Quantitative Sampling and Structure-Aware Attention

引用

IEEE ACCESS 2024年 12卷 25512-25524页

作者： Zhao, Qin Miao, Yaru An, Dongdong Lian, Jie Li, Maozhen Shanghai Normal Univ Shanghai Engn Res Ctr Intelligent Educ & Big Data Shanghai 200234 Peoples R China Tongji Univ Key Lab Embedded Syst & Serv Comp Minist Educ Shanghai 200092 Peoples R China Brunel Univ London Dept Elect & Comp Engn London UB8 3PH England

Heterogeneous information networks provide abundant structural and semantic information. Two main strategies for leveraging this data include meta-path-based and meta-path-free methods. The effectiveness of the former heavily depends on the quality of manually defined meta-paths, which may lead to the instability of the model. However, the existing meta-path-free methods lack of neighbor screening during aggregating, and there is also an overemphasis on attribute information. To address these issues, we propose the Heterogeneous Graph Neural Network model by incorporating Quantitative Sampling and Structure-aware Attention. We introduce a Quantitative Sampling Module that calculates the similarity between neighbors of the target nodes and target nodes, enabling us to select the top k nodes with the strongest relevance to the target node based on this measure, and incorporate a Structure-aware Attention Module during the aggregation of neighbor information. This module combines both structural and attribute information to aggregate the neighbor information effectively. By implementing these improvements, our proposed model exhibits superior performance compared to several state-of-the-art methods on two real-world datasets.

关键词： Graph neural networks Social networking (online) Correlation Prediction algorithms Task analysis Stacking Semantics Information systems Heterogeneous networks detection algorithms Heterogeneous information network community detection quantitative sampling structure-aware attention

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Analysis of the Machine Learning algorithms in IoT IDS Systems

引用

IEEE ACCESS 2024年 12卷 46785-46811页

作者： Ozdogan, Erdal Gazi Univ IT Dept TR-06560 Ankara Turkiye

In this study, machine learning algorithms in IoT IDS (Internet of Things Intrusion detection System) systems are comprehensively compared from various aspects. Accuracy, precision, and training time are evaluated. The effects of data preprocessing techniques including normalization, outlier removal, standardization, and regularization on the datasets are examined. Furthermore, the impact of dataset balancing, considering both balanced and imbalanced scenarios, on machine learning performance is investigated. The contribution of feature selection on the four different datasets is also analyzed. Based on findings, it is observed that certain preprocessing operations provide significant advantages in various ML algorithms, whereas others have very low impact, and their performance varies depending on the dataset and feature selection. The aim of this study is to facilitate the complexity and lengthiness of machine learning processes and algorithm selection, providing insights for future academic research. By addressing this objective, an effort is made to shed light on simplifying the utilization of machine learning algorithms. The challenges arising from the complexity of machine learning processes in IoT IDS systems are addressed by this study. This contribution can greatly benefit researchers in their academic endeavors. This multifaceted approach proves beneficial when comparing the methods under consideration, fostering a scientific discourse on their efficacy within contexts.

关键词： detection algorithms IoT attack detection IoT IDS ML algorithm comparison machine learning preprocessing machine learning in IoT IDS

来源：评论

学校读者我要写书评

暂无评论

Label Propagation Algorithm Joint Multilayer Neighborhood Overlap and Historic Label Similarity for Community detection

引用

IEEE SYSTEMS JOURNAL 2022年第2期16卷 2626-2634页

作者： Zhang, Siyang Yu, Siyu E, Xinhua Huo, Ru Sui, Ziheng Beijing Univ Posts & Telecommun Beijing 100876 Peoples R China Beijing Univ Technol Fac Informat Technol Beijing 100124 Peoples R China Peking Univ Fac Informat Technol Sch Software & Microelect Beijing 102600 Peoples R China

Community detection is an extremely important technology for today's rapidly evolving data mining and exploratory analysis. Many community detection algorithms have been proposed and applied. Among them, the label propagation algorithm (LPA) is widely used due to its closeness to linear time complexity and its simple characteristics. However, the results of detecting the community were found to be random when using LPA. Based on previous improvements to the LPA, this article proposes a new LPA (NOHLPA) that joins multilayer neighborhood overlap and historical label similarity. The NOHLPA considers both the node update order and label selection rules. We cited the label entropy as the basis for node update order and defined multilayer neighborhood overlap and historical label similarity for calculating node preferences to devise more appropriate label selection rules. We test the NOHLPA on five real datasets and compare NOHLPA with different combinations of algorithms. The experimental results show that the NOHLPA effectively improves the accuracy of community partitioning while ensuring stability.

关键词： Entropy Nonhomogeneous media Internet Time complexity Social networking (online) Heuristic algorithms detection algorithms Community detection historical label similarity label propagation multilayer neighborhood overlap

来源：评论

学校读者我要写书评

暂无评论

Infrared Small UAV Target detection Algorithm Based on Enhanced Adaptive Feature Pyramid Networks

引用

IEEE ACCESS 2022年 10卷 115988-115995页

作者： Zheng Lu Peng Yueping Ye Zecong Jiang Rongqi Zhou Tongtong Engn Univ PAP Xian 710086 Peoples R China Engn Univ PAP Coll Informat Engn Xian 710086 Peoples R China

Under the explosive growth of "low-slow-small" unmanned aerial vehicles (UAVs), harassment incidents with "low-slow-small" UAVs, such as illegal surveying and mapping, no-fly zone flights, terrorist attacks, have frequently been occurring due to the low entry barrier to the industry and the lack of market norms and control standards, and severely impair the safety of public life and property. The premise of developing anti-UAV technology is efficient and accurate UAV target detection, which plays a vital role in the follow-up UAV target tracking, attack, and interception. This paper proposes a new DEAX algorithm named enhanced adaptive feature pyramid networks-based Target detection Algorithm for infrared small UAV target detection. Our proposed algorithm improves the original feature pyramid networks in four aspects for the small UAV target detection task. 1) Use channel separation to keep more channels when adjusting the number of channels in convolutional to avoid the loss of helpful feature information;2) Design a feature enhancement module to enhance "target feature" and suppress "non-target feature";3) Alleviate the differences of receptive fields and semantic information among different layers by shared convolution;4) Introduce adaptive feature fusion method into feature fusion and complete the construction of enhanced adaptive feature pyramid networks (EAFPN) to solve the problem of weakening feature expression in cross-scale feature fusion. EAFPN is applied to the single-stage target detection networks. We find the detection accuracy and speed of the algorithm outperform those based on feature pyramid networks (FPN), where the improvement on mAP is 8.9%.

关键词： Feature extraction Object detection Semantics detection algorithms Target tracking Imaging Classification algorithms Object detection feature pyramid feature enhancement

来源：评论

学校读者我要写书评

暂无评论

A Methodology for Forest Fire detection and Notification Using AI and IoT Approaches 2

A Methodology for Forest Fire Detection and Notification Usi...

引用

2nd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2024

作者： Borah, Asha Rani Ravi, Nikhil Narayan, Nikhil D'Souza, Prathik Joel New Horizon College of Engineering Computer Science and Engineering Bengaluru India

ISBN: (纸本)9798350327533

This study explores how advanced technology has revolutionized the use of drones for forest fire surveillance. This research study explores the historical development of Unmanned Aerial Vehicles (UAVs) in forest monitoring with a focus on control systems and detection algorithms. The drone's capabilities are improved via telemetry integration, sophisticated autopilot systems, and an emphasis on accuracy, autonomy, and real-Time communication. An analysis of object recognition, anomaly detection, and environmental monitoring algorithms in detail demonstrates how important a role these algorithms play in spotting dangers like forest fire. This study presents successful implementations through illustrative case studies, discusses obstacles, and lays forth future plans for utilizing drone technology for the forest fire surveillance in environmental conservation initiatives. © 2024 IEEE.

关键词： Autonomy Control systems Cutting-edge technology detection algorithms Forest surveillance drones Historical evolution Real-Time communication Unmanned Aerial Vehicles

来源：评论

学校读者我要写书评

暂无评论

Improving Plant Disease Classification With Deep-Learning-Based Prediction Model Using Explainable Artificial Intelligence

引用

IEEE ACCESS 2024年 12卷 100005-100014页

作者： Nigar, Natasha Faisal, Hafiz Muhammad Umer, Muhammad Oki, Olukayode Manappattukunnel Lukose, Jose Univ Engn & Technol Dept Comp Sci RCET Lahore 54000 Pakistan Tech Univ Munich Sch Engn & Design D-80333 Munich Germany TEKHQS INC Irvine CA 92620 USA Walter Sisulu Univ Dept Informat Technol ZA-5117 Mthatha South Africa

Plant diseases can have profound effects on the economy, impacting both local and global scales. These diseases can lead to substantial losses in agricultural productivity, affecting crop yields and quality. In this context, deep learning algorithms are widely acknowledged as effective solutions. However, the use of these black-box approaches raises concerns about trust in interpreting and validating the decisions generated by the models. This study proposes an explainable artificial intelligence (XAI) based plant disease classification system to classify and identify distinct ailments with improved accuracy. The system correctly identifies 38 different plant diseases with accuracy, precision, and recall as 99.69%, 98.27%, and 98.26%, respectively. These predictions are subjected to additional analysis employing the local interpretable model-agnostic explanations (LIME) framework to produce visual explanations aligning with prior beliefs and adhering to established best practices in explanations. This system will serve as a promising avenue for revolutionizing disease detection, fostering informed decision-making, and ultimately contributing to global food security.

关键词： Accuracy Plant diseases Feature extraction Solid modeling Computational modeling Predictive models Explainable AI detection algorithms Plant disease detection deep learning explainable artificial intelligence prediction model prediction model

来源：评论

学校读者我要写书评

暂无评论

AdamW plus : Machine Learning Framework to Detect Domain Generation algorithms for Malware

引用

IEEE ACCESS 2024年 12卷 79138-79150页

作者： Javed, Awais Rashid, Imran Tahir, Shahzaib Saeed, Saqib Almuhaideb, Abdullah M. Alissa, Khalid Natl Univ Sci & Technol NUST Coll Signals Islamabad 44000 Pakistan Imam Abdulrahman Bin Faisal Univ Coll Comp Sci & Informat Technol Saudi Aramco Cybersecur Chair Dept Comp Informat Syst Dammam 31441 Saudi Arabia Imam Abdulrahman Bin Faisal Univ Coll Comp Sci & Informat Technol Saudi Aramco Cybersecur Chair Dept Networks & Commun Dammam 31441 Saudi Arabia

Advanced Persistent Threats commonly use Domain Generation algorithms to evade advanced detection methods to establish communication with their command and control servers. To overcome the DNS protocol legitimacy breach, several Domain Generation algorithms detection methods have been proposed. To solve the problem of DGA malware detection amicably, Deep Learning based detection schemes attracted researchers' interest recently. However, Deep Learning has already achieved optimal results and the gap is identified as fine-tuning of Deep Learning model hyper-parameters. The proposed solution is focusing on a model specific hyper-parameter known as the gradient optimizer. Gradient Optimisers are broadly categorised into Stochastic Gradient and Adaptive Moment based Gradient. Moment-based Gradient optimizer approaches are identified with suffering from weight decay and leading to poor generalization. Adaptive Moment (Adam) has improved with weight Decay as AdamW. To optimise moment-based gradient optimizers, Adam and AdamW are analyzed deeply. To optimize the functioning of AdamW, we present AdamW+, a novel solution for detecting DGA algorithms through re-implementing and nullifying the weight decay in AdamW. AdamW+ has been successfully implemented and shown promising results compared to Adam and AdamW optimizers in practice. AdamW+ preserved the properties of Adaptive Optimizer Adam while simplifying the weight decay implementation of AdamW. Empirical analysis has proved that AdamW+ has outperformed Adam and AdamW. The experimental result have substantiated that the proposed algorithm achieves the best accuracy result.

关键词： Deep learning Malware Long short term memory Servers Protocols Domain Name System Arithmetic Optimization methods Threat assessment detection algorithms Stochastic processes Command and control systems Deep learning optimizer LSTM Adam AdamW AdamW plus

来源：评论

学校读者我要写书评

暂无评论

MMSRC: A Multidirection Multiscale Spectral-Spatial Residual Network for Hyperspectral Multiclass Change detection

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2022年 15卷 9254-9265页

作者： Ge, Hongmei Tang, Yongsheng Bi, Zuolin Zhan, Tianming Xu, Yang Song, Aibo Southeast Univ Sch Comp Sci & Engn Nanjing 211189 Peoples R China Nanjing Audit Univ Sch Engn Audit Nanjing 211815 Peoples R China Nanjing Audit Univ Sch Informat Engn Nanjing 211815 Peoples R China Nanjing Audit Univ Jiangsu Key Construct Lab Audit Informat Engn Nanjing 211815 Peoples R China Nanjing Univ Sci & Technol Nanjing 210094 Peoples R China Southeast Univ Sch Comp Sci & Engn Nanjing 211189 Peoples R China

Recently, deep convolutional neural network (CNN) hyperspectral change detection methods have achieved significant improvement. However, most CNN hyperspectral change detection methods do not make full use of spectral-spatial feature information. In this article, we propose a novel multidirection and multiscale spectral-spatial residual network for hyperspectral multiclass change detection. Specifically, a multiscale structure and a multidirection mechanism are introduced to investigate feature variation of hyperspectral images and improve the accuracy of hyperspectral change detection. Experiments on multiple hyperspectral datasets show that the proposed method achieves improved performance in comparison with other advanced hyperspectral multiclass change detection methods.

关键词： Hyperspectral imaging Feature extraction Convolution Tensors detection algorithms Deep learning Residual neural networks Convolutional neural network (CNN) hyperspectral change detection multidirection multiscale

来源：评论

学校读者我要写书评

暂无评论

Density-Peak-Based Overlapping Community detection Algorithm

引用

IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS 2022年第4期9卷 1211-1223页

作者： Sun, Liping Ye, Tao Sun, Jian Duan, Xiaoyu Luo, Yonglong Anhui Normal Univ Sch Comp & Informat Anhui Prov Key Lab Network & Informat Secur Wuhu 241003 Peoples R China Nanjing Univ Software Inst Nanjing 210023 Peoples R China

Overlapping community detection is essential for revealing the hidden structure of complex networks. In this work, we present an overlapping community detection algorithm that selects community centers adaptively based on density peaks. The proposed algorithm, called the density-peak-based overlapping community detection (DPOCD) algorithm, defines point link strength and edge link strength to construct distance matrix. Unlike the density peaks clustering algorithm, by which cluster centers are selected manually, the DPOCD algorithm uses the linear fitting method to select community centers. To evaluate the feasibility of the presented algorithm, we compared it with other advanced methods on artificial synthetic network and real complex network datasets. The experimental results demonstrate that our method achieves excellent performance in large-scale complex networks and the robustness of the algorithm.

关键词： Complex networks Clustering algorithms detection algorithms Sun Optimization Image edge detection Robustness Adaptive selection of cluster centers community detection algorithm density peak overlapping community

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：