检索结果-内蒙古大学图书馆

Video and image quality improvement using an enhanced optimized dehazing technique

Multimedia Tools and Applications 2024年 1-19页

作者： Ayoub, Abeer El-Shafai, Walid El-Samie, Fathi E. Abd Hamad, Ehab K. I. EL-Rabaie, El-Sayed M. Electrical Engineering Department Faculty of Engineering Aswan University Aswan81542 Egypt Department of Electronics and Electrical Communication Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt Security Engineering Lab Computer Science Department Prince Sultan University Riyadh11586 Saudi Arabia Department of Information Technology College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia

The detrimental effects of atmospheric haze frequently plague outdoor imagery. This phenomenon arises from the scattering of light by minute particles within the ambient environment surrounding the scene to be imaged. Haze engenders an overall whitening of the image, leading to diminished contrast. To address these issues and enhance the quality of hazy images and videos, an enhanced dehazing technique is proposed. The proposed technique includes image enhancement before the optimal dehazing process. The enhancement stage entails the implementation of both homomorphic processing and Contrast Limited Adaptive Histogram Equalization (CLAHE), serving to control the image dynamic range, while concurrently heightening the image contrast. The culminating stage encompasses an optimized dehazing technique, adept at expunging haze-induced artifacts from images. The homomorphic processing and CLAHE, applied in the pre-processing step, establish a foundation for the subsequent dehazing procedure. This proposed methodology is systematically applied to a gamut of visual outputs, including visible videos, Near-Infrared (NIR) frames, and authentic hazy images. Comparative evaluations of the proposed technique, homomorphic-processing-based enhanced dehazing, and standalone dehazing techniques is undertaken on different video types encompassing five frames. It is evident that the proposed technique, synergizing the homomorphic processing, CLAHE, and dehazing, outperforms alternative strategies. Furthermore, the proposed technique is subjected to a comparative study with various existing dehazing techniques over real hazy images. The assessment depends on Peak Signal-to-Noise Ratio (PSNR), correlation, and entropy metrics. The results underscore the efficacy of the proposed technique, particularly in terms of spectral entropy enhancement of dehazed frames. For both visible and NIR frames, the percentages of enhancement by the proposed technique are 17.66% and 118.48%, respectively. ©

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

A variable structure model of neurons’ application in computer vision

引用

Multimedia Tools and Applications 2025年 1-11页

作者： Sridi, Alaeddine Attia, Mouna Bouallegue, Kais Electronics and Micro-Electronics Laboratory Faculty of Sciences of Monastir University of Monastir Monastir Tunisia The Higher Institute of Computer Science and Communication Technologies of Hammam Sousse University of Sousse Sousse Tunisia Laboratory of Computer Science Parallelism and Intelligent Communicating Systems Faculty of Sciences of Tunisia University of Tunis ELManar Tunis Tunisia Department of Electrical Engineering Higher Institute of Applied Sciences and Technology of Sousse University of Sousse Sousse Tunisia

The neuron doctrine defines the neuron as the basic unit of the nervous system, which drives the dynamic behavior of our organs. This has led to neurons becoming the focus of modern neuroscience research and to the rise of neurocomputing, in which we try to digitally simulate human behavior. The goal being to better understand the human nervous system and gain more control over neuron-related diseases, helping doctors establish faster and more reliable medical diagnoses. In this context, we aim to simulate eye diseases and vision problems by convolving input images using a new class of neural networks based on a variable structure model of neurons presented in [1]. In this work, we focus on recreating the common convolution filters as a first step. This is the first time in neurocomputing literature that dendritic neurons have been used for dynamic convolution kernels generation and eye diseases simulation. We start with a presentation of our upgrades to the model’s flexibility and accuracy. Then, we propose a novel kernel generator algorithm to convolve the input images. Finally, we show its promising results in creating the outline, pixelization, and night convolution kernels, as well as control some metrics, including brightness and data retention. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

Electricity Consumption Classification using Various Machine Learning Models

EAI Endorsed Transactions on Energy Web

引用

EAI Endorsed Transactions on Energy Web 2024年 11卷 1-6页

作者： Paikaray, Bijay Kumar Jena, Swarna Prabha Mondal, Jayanta Van Thuan, Nguyen Tung, Nguyen Trong Mallick, Chandrakant Bhubaneswar India Department of Electronics and Communication Engineering Centurion of University of Technology and Management Odisha India School of Computer Engineering KIIT Deemed to be University Bhubaneswar India Faculty of Engineering Technology Hung Vuong University Viet Nam Quality Assurance Dong A University Viet Nam Department of Computer Science & Information Technology GITA Autonomous College Odisha Bhubaneswar India

INTRODUCTION: As population has increased over successive generations, human dependency on electricity has increased to the point where it has become a norm and indispensable, and the idea of living without it has become unthinkable. OBJECTIVES: Machine learning is emerging as a fundamental method for performing tasks autonomously without human intervention. Forecasting electricity consumption is challenging due to the many factors that influence it;embracing modern technology with its heavy focus on machine learning and artificial intelligence is a potential solution. METHODS: This study employs various machine learning algorithms to forecast power usage and determine which method performs best in predicting the dataset based on different variables. RESULTS: Eight models were tested, including Linear Regression, DT Classifier, RF Classifier, KNN, DT Regression, SVM, Logistic Regression, and GNB Classifier. The Decision Tree model had the greatest accuracy of 98.3%. CONCLUSION: The Decision Tree model’s accuracy can facilitate efficient use of electricity, leading to both conservation of electricity and cost savings, and be a guiding light in future planning. Copyright © 2024 Paikaray, B.K. et al., licensed to EAI. This open-access article is distributed under the terms of the CC BY-NC-SA 4.0, which permits copying, redistributing, remixing, transforming, and building upon the material in any medium so long as the original work is properly cited.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Spatial Attention Integrated EfficientNet Architecture for Breast Cancer Classification with Explainable AI

引用

computers, Materials & Continua 2024年第9期80卷 5029-5045页

作者： Sannasi Chakravarthy Bharanidharan Nagarajan Surbhi Bhatia Khan Vinoth Kumar Venkatesan Mahesh Thyluru Ramakrishna Ahlam AlMusharraf Khursheed Aurungzeb Department of Electronics and Communication Engineering Bannari Amman Institute of TechnologySathyamangalam638402India School of Computer Science Engineering and Information Systems(SCORE) Vellore Institute of TechnologyVellore632014India School of Science Engineering and EnvironmentUniversity of SalfordManchesterM54WTUK Department of Computer Science&Engineering Faculty of Engineering and TechnologyJAIN(Deemed-to-be University)Bengaluru562112India Department of Management College of Business AdministrationPrincess Nourah Bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia Department of Computer Engineering College of Computer and Information SciencesKing Saud UniversityP.O.Box 51178Riyadh11543Saudi Arabia Adjunct Research Faculty Centre for Research Impact&OutcomeChitkara UniversityRajpura140401India

Breast cancer is a type of cancer responsible for higher mortality rates among *** cruelty of breast cancer always requires a promising approach for its earlier *** light of this,the proposed research leverages the representation ability of pretrained EfficientNet-B0 model and the classification ability of the XGBoost model for the binary classification of breast *** addition,the above transfer learning model is modified in such a way that it will focus more on tumor cells in the input ***,the work proposed an EfficientNet-B0 having a Spatial Attention Layer with XGBoost(ESA-XGBNet)for binary classification of *** this,the work is trained,tested,and validated using original and augmented mammogram images of three public datasets namely CBIS-DDSM,INbreast,and MIAS *** accuracy of 97.585%(CBISDDSM),98.255%(INbreast),and 98.91%(MIAS)is obtained using the proposed ESA-XGBNet architecture as compared with the existing ***,the decision-making of the proposed ESA-XGBNet architecture is visualized and validated using the Attention Guided GradCAM-based Explainable AI technique.

关键词： EfficientNet mammograms breast cancer Explainable AI deep-learning transfer learning

来源：评论

学校读者我要写书评

暂无评论

Smart Drive Safe: Harnessing CNN for Enhanced Traffic Sign Recognition and Voice Alerts 5

Smart Drive Safe: Harnessing CNN for Enhanced Traffic Sign R...

引用

5th International Conference on Electronics and Sustainable communication Systems, ICESC 2024

作者： Priya, D. Divya Chinnasamy, P. Ayyasamy, Ramesh Kumar Kiran, Ajmeera Jalil, Norazira Binti A. Sangodiah, Anbuselvan Mlr Institue of Technology Department of Computer Science Engineering Hyderabad India School of Computing Kalasalingam Academy of Research and Education Department of Computer Science and Engineering Tamil Nadu Srivilliputtur India Faculty of Information and Communication Technology Perak Kampar Malaysia Quest International University Faculty of Computing and Engineering Perak Ipoh Malaysia

ISBN: (纸本)9798350379945

Traffic signs secure their safety on the roads and the protection of others around them;everyone must be aware of traffic signs and other aspects of road safety. Traffic sign detection is a highly relevant computer vision problem that forms the basis of numerous applications in sectors like the automotive industry. Between classes, traffic signs differ widely in terms of colour, shape, and the use of pictograms or text. The project's classification of traffic signs in an image into various categories is based on a deep neural network model. IoT devices that capture traffic signs and notify users of them are used to build a model. © 2024 IEEE.

关键词： Automotive industry

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Keylogger Detection Systems Using Recurrent Neural Networks Enabled with Feature Selection Model 7th

An Enhanced Keylogger Detection Systems Using Recurrent Neur...

引用

7th International Conference on communication, Device and Networking, ICCDN 2024

作者： Awotunde, Joseph Bamidele Sur, Samarendra Nath Imoize, Agbotiname Lucky Rodríguez, Demóstenes Zegarra Akanji, Boluwatife Department of Computer Science Faculty of Information and Communication Sciences University of Ilorin Kwara State Ilorin240003 Nigeria Department of Electronics and Communication Engineering Sikkim Manipal Institute of Technology Sikkim Manipal University Sikkim India Department of Electrical and Electronics Engineering Faculty of Engineering University of Lagos Akoka Lagos 100213 Nigeria Department of Electrical Engineering and Information Technology Institute of Digital Communication Ruhr University Bochum44801 Germany Department of Computer Science Federal University of Lavras Minas Gerais Lavras Brazil

ISBN: (纸本)9789819764648

Keyloggers are malicious software programs that record keystrokes of users without their consent or knowledge. They can steal sensitive information like credit card numbers and passwords. They pose a significant threat to users’ privacy and security since they capture keystroke data. Keylogger detection systems play a vital role in safeguarding users from cyberattacks and mitigating the potential harm caused by keyloggers. Keylogger detection systems employing deep learning have been widely used for identifying and mitigating cyber threats associated with keyloggers. However, they struggle to identify new or unfamiliar keylogger samples. This study aims to explore deep learning techniques with feature selection model to detect keylogger. This study employs Recurrent Neural Networks (RNNs) using a collection of known keylogger dataset. Furthermore, a correlation-based feature extraction method was applied to identify the most relevant features for the model, highlighting the importance of specific features, such as URG flag Count, ACK flag count, and Idle Mean, in differentiating between benign and keylog classes. The proposed model demonstrates superior accuracy of 0.8763 and precision of 0.8569 compared to a baseline model using Logistic Regression (LR) 0.7721 and 0.7407, respectively, indicating a better balance between accurate classification and minimizing false negatives and false positives. The findings from this study help to improve keylogger detection methods, making them more powerful and efficient. These results suggest that a keylogger-based detection system employing a deep learning approach can be a valuable tool for addressing the complex and evolving landscape of cyber threats related to keylogging. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

An eBPF-Based Trace-Driven Emulation Method for Satellite Networks

IEEE Networking Letters

引用

IEEE Networking Letters 2024年第3期6卷 188-192页

作者： Tian, Weibiao Li, Ye Zhao, Jinwei Wu, Sheng Pan, Jianping Nantong University School of Information Science and Technology Nantong226019 China Nantong Research Institute for Advanced Communication Technologies Nantong226019 China Beijing University of Posts and Telecommunications School of Information and Communication Engineering Beijing100876 China University of Victoria Faculty of Engineering and Computer Science VictoriaBCV8P 5C2 Canada

System-level performance evaluation over satellite networks often requires a simulated or emulated environment for reproducibility and low cost. However, the existing tools may not meet the needs for scenarios such as the low-earth orbit (LEO) satellite networks. To address the problem, this letter proposes and implements a trace-driven emulation method based on Linux's eBPF technology. Building a Starlink traces collection system, we demonstrate that the method can effectively and efficiently emulate the connection conditions, and therefore provides a means for evaluating applications on local hosts. © 2019 IEEE.

关键词： Satellite communication systems

来源：评论

学校读者我要写书评

暂无评论

Simple ARQ Protocol for Reliable Transport in LowPANs 21

Simple ARQ Protocol for Reliable Transport in LowPANs

引用

2024 IEEE/ACS 21st International Conference on computer Systems and Applications, AICCSA 2024

作者： Chahed, Manel Meddeb, Aref Boufaied, Amine Higher Institute of Computer Science and Communication Technologies of Sousse University of Sousse NOCCS Research Laboratory Sousse Tunisia University of Sherbrooke Faculty of Engineering Sherbrooke Canada Higher Institute of Computer Science and Communication Technologies of Sousse University of Sousse MARS Research Laboratory Sousse Tunisia

ISBN: (纸本)9798331518240

Due to the surge in IoT devices, numerous protocols have been proposed to meet their needs. UDP is commonly used for IoT because of its simplicity, low latency, minimal overhead, and low energy consumption. In contrast, TCP is less suitable for IoT due to its higher resource demands, complexity, and greater energy consumption, which are challenging for small devices with limited resources. Therefore, often, reliability is provided by upper-layer protocols, mainly by the applications themselves. In fact, classical IoT applications such as sensing, identification and actuating generate multiple copies of data due to the hardware redundancy and periodic updates. Examples of such applications include, agricultural, environmental, traffic, and healthcare monitoring. UDP and TCP protocols may be inadequate for IoT applications requiring reliable, real-time communication. In critical situations like battlefields or disasters, sensors might only send a few messages before being destroyed. Therefore, the network itself must ensure reliability in these scenarios. Moreover, smart textiles are more and more integrating sensory devices and require reliable transmissions. In fact, with the very stringent resource constraints on one hand, and the requirements to respect the Specific Absorption Rate (SAR) of human bodies, on the other hand, re-transmissions and power must be kept at their lowest levels. In this paper, we propose a simple Automatic Request (SARQ) transport protocol that uses acknowledgments and a retransmission mechanism. Through a realistic simulation setup using Contiki motes in cooja simulator, we show that our protocol exhibits slightly higher energy consumption and resource requirements than UDP, but far less than TCP. Conversely, we show that our protocol exhibits 99% Packet Delivery Ratio (PDR), while UDP and TCP exhibit 74% and 99% PDR, respectively. © 2024 IEEE.

关键词： Smart textiles

来源：评论

学校读者我要写书评

暂无评论

Effective Return Rate Prediction of Blockchain Financial Products Using Machine Learning

引用

computers, Materials & Continua 2023年第1期74卷 2303-2316页

作者： K.Kalyani Velmurugan Subbiah Parvathy Hikmat A.M.Abdeljaber T.Satyanarayana Murthy Srijana Acharya Gyanendra Prasad Joshi Sung Won Kim PG and Research Department of Computer Science Marudupandiyar College(Affiliated to Bharathidasan UniversityTiruchirappalli)Thanjavur613403India Department of Electronics and Communication Engineering Kalasalingam Academy of Research and EducationKrishnankoil626126TamilnaduIndia Department of Computer Science Faculty of Information TechnologyApplied Science Private UniversityAmmanJordan Department of Information Technology Chaitanya Bharathi Institute of TechnologyHyderabadTelangana500075India Department of Convergence Science Kongju National UniversityGongju32588Korea Department of Computer Science and Engineering Sejong UniversitySeoul05006Korea Department of Information and Communication Engineering Yeungnam UniversityGyeongsan-siGyeongbuk-do38541Korea

In recent times,financial globalization has drastically increased in different ways to improve the quality of services with advanced *** successful applications of bitcoin Blockchain(BC)techniques enable the stockholders to worry about the return and risk of financial *** stockholders focused on the prediction of return rate and risk rate of financial ***,an automatic return rate bitcoin prediction model becomes essential for BC financial *** newly designed machine learning(ML)and deep learning(DL)approaches pave the way for return rate predictive *** study introduces a novel Jellyfish search optimization based extreme learning machine with autoencoder(JSO-ELMAE)for return rate prediction of BC financial *** presented JSO-ELMAE model designs a new ELMAE model for predicting the return rate of financial ***,the JSO algorithm is exploited to tune the parameters related to the ELMAE model which in turn boosts the classification *** application of JSO technique assists in optimal parameter adjustment of the ELMAE model to predict the bitcoin return *** experimental validation of the JSO-ELMAE model was executed and the outcomes are inspected in many *** experimental values demonstrated the enhanced performance of the JSO-ELMAE model over recent state of art approaches with minimal RMSE of 0.1562.

关键词： Financial products blockchain return rate prediction model machine learning parameter optimization

来源：评论

学校读者我要写书评

暂无评论

NEAR: Non-Supervised Explainability Architecture for Accurate Review-Based Collaborative Filtering

引用

IEEE Transactions on Knowledge and Data engineering 2024年第2期36卷 750-765页

作者： Pugoy, Reinald Adrian Kao, Hung-Yu National Cheng Kung University Intelligent Knowledge Management Lab Department of Computer Science and Information Engineering Tainan City701 Taiwan University of the Philippines Open University Faculty of Information and Communication Studies Los Baños4031 Philippines

There is a critical issue in explainable recommender systems that compounds the challenges of explainability yet is rarely tackled: the lack of ground-truth explanation texts for training. It is unrealistic to expect every user-item pair in a dataset to have a corresponding target explanation. Hence, we pioneer the first non-supervised explainability architecture for review-based collaborative filtering (called NEAR) as our novel contribution to the theory of explanation construction in recommender systems. While maintaining excellent recommendation performance, our approach reformulates explainability as a non-supervised (i.e., unsupervised and self-supervised) explanation generation task. We formally define two explanation types, both of which NEAR can produce. An invariant explanation, fixed for all users, is based on the unsupervised extractive summary of an item's reviews via embedding clustering. Meanwhile, a variant explanation, personalized for a specific user, is a sentence-level text generated by our customized Transformer conditioned on every user-item-rating tuple and artificial ground-truth (self-supervised label) from one of the invariant explanation's sentences. Our empirical evaluation illustrates that NEAR's rating prediction accuracy is better than the other state-of-the-art baselines. Moreover, experiments and assessments show that NEAR-generated variant explanations are more personalized and distinct than those from other Transformer-based models, and our invariant explanations are preferred over those from other contemporary models in real life. © 1989-2012 IEEE.

关键词： Recommender systems

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：