检索结果-内蒙古大学图书馆

Video quality enhancement using different enhancement and dehazing techniques

Journal of Ambient Intelligence and Humanized Computing 2023年第12期14卷 16607-16635页

作者： Ayoub, Abeer Naeem, Ensherah A. El-Shafai, Walid El-Samie, Fathi E. Abd Hamad, Ehab K. I. El-Rabaie, El-Sayed M. Department of Electronics and Electrical Communications Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt Electrical Department Faculty of Technology and Education Suez University Suez43527 Egypt Security Engineering Lab Computer Science Department Prince Sultan University Riyadh11586 Saudi Arabia Electrical Engineering Department Faculty of Engineering Aswan University Aswan81542 Egypt Department of Information Technology College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia

Video in an outdoor environment suffers from the haze problem due to smoke, dust, and other particles in the atmosphere. The quality of video under these atmospheric conditions is clearly/dramatically degraded, leading to poor contrast and information leakage. The proposed algorithm in this paper depends on pre-processing of video frames prior to the dehazing process. Different enhancement techniques are used to remove noise and control the dynamic range, because each frame has some noise due to sensor measurement errors. This noise might be magnified during the haze removal process, if totally neglected. Previous dehazing algorithms enhance the video contrast and reduce frame degradation. We use optimized dehazing algorithms, namely a dual-transmission-map dehazing algorithm, and a recursive Deep Residual Learning (DRL) network, as dehazing tools. We modify the previous dehazing algorithms to generate new algorithms to be suitable not only for visible frames but also for Near Infrared (NIR) frames. In the optimized dehazing algorithms, the haze effect is reduced, according to the hazing model. In the dual-transmission-map dehazing algorithm, the effect of attenuation parameter R and attenuation weight U on the dehazed frames is investigated. A dual-transmission-map algorithm, depending on the Dark Channel Prior (DCP) technique, is studied. We study the effect of a regularization parameter Ω on the visible and NIR dehazed frames without and with enhancement techniques. For the recursive deep residual learning algorithm, we increase the number of iterations in the DRL network from three iterations in the traditional DRL algorithm without denoising techniques to nine iterations, in order to explore the impact of increasing the number of iterations on the output dehazed frames. Since elapsed time grows with the number of iterations, we terminate the operation after the ninth iteration. A comparison between traditional algorithms and the proposed dehazing algorithms is

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Predictive Modeling of Chronic Kidney Disease Progression Using Longitudinal Clinical Data and Deep Learning Techniques 2

Predictive Modeling of Chronic Kidney Disease Progression Us...

引用

2nd International engineering Conference on Electrical, Energy, and Artificial Intelligence, EICEEAI 2023

作者： Rastegari, Farzaneh Odeh, Mahmoud Baroughi, Reyhaneh Mehdizadeh Bustan, Arad University of Connecticut Computer Science and Engineering Department Storrs United States School of Information Technology Zarqa University Cyber Security Department Zarqa Jordan College of Business Administration University of Business and Technolog Jeddah21448 Saudi Arabia Electrical and Computer Engineering Polytechnic University of Turin Turin Italy University of Maine Orono United States

ISBN: (纸本)9798350373363

Chronic Kidney Disease (CKD) is a prevalent condition with significant morbidity and mortality. Early and accurate detection is critical for effective treatment and management. This paper presents a comparative analysis of deep learning techniques, specifically Multi-layer Perceptron (MLP) models, for predicting CKD. We preprocessed a dataset containing clinical and demographic features, trained MLP models of varying complexities, and evaluated their performances. The results showed that simpler models achieved high accuracy, precision, recall, and F1 scores, comparable to more complex models. These findings suggest that deep learning can be a valuable tool in the early detection of CKD, with the potential to support clinical decision-making. © 2023 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Algorithmic Innovations in Machine Learning for Drug Discovery and Pharmaceutical Innovations 2

Algorithmic Innovations in Machine Learning for Drug Discove...

引用

2nd IEEE International Conference on Trends in Quantum Computing and Emerging Business Technologies, TQCEBT 2024

作者： Sangeetha, M. Harin Fernandez, F. Mary Ganesh Ramachandran, A. Saravanan, S.K. Bhanumathi, M. University of Technology and Applied Sciences Department of Computer Engineering Nizwa Oman Saveetha Institute of Medical and Technical Sciences Saveetha University Saveetha School of Engineering Department of Computer Science and Engineering Chennai India Saveetha Institute of Medical and Technical Sciences Saveetha University Saveetha School of Engineering Department of Computer Science Chennai India Saveetha School of Engineering Simats BigData and Network Security Department of Computers Science and Engineering Tamil Nadu Chennai India Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350384277

Algorithmic advancements in machine learning have transformed drug discovery and pharmaceutical innovation. This study examines how machine learning might speed medication development and pharmaceutical innovation. Data from genomes, chemistry, and clinical trials is collected to start the research. Careful data cleansing, normalization, and feature engineering prepare the data for model building. Random Forest, SVM, NN, Decision Tree, and Logistic Regression are tested for prediction. ROC AUC results show that Random Forest is the most accurate model at 86%. Ensemble approaches have drug discovery potential. Beyond accuracy, models are assessed for precision, false positive, and false negative rates. Drug discovery relies on precision to measure positive predictions. The Random Forest model makes the most accurate positive predictions at 0.87. However, false positives and negatives also matter in drug discovery. The Decision Tree model contains the falsest positives (35), underlining the need to reduce them to save resources. The Neural Network approach reduces false negatives to 12, ensuring good medication candidates are not overlooked. The results show that different machine learning models have trade-offs, underscoring the necessity for a personalized strategy depending on objectives and restrictions. Algorithmic breakthroughs in pharmaceutical innovation reduce medication development time by 40% and cost by 25%, according to the study. © 2024 IEEE.

关键词： Drug discovery

来源：评论

学校读者我要写书评

暂无评论

On the Effectiveness of Function-Level Vulnerability Detectors for Inter-Procedural Vulnerabilities 24

On the Effectiveness of Function-Level Vulnerability Detecto...

引用

44th ACM/IEEE International Conference on Software engineering, ICSE 2024

作者： Li, Zhen Wang, Ning Zou, Deqing Li, Yating Zhang, Ruqian Xu, Shouhuai Zhang, Chao Jin, Hai Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security Cluster and Grid Computing Lab National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hong Kong Jin YinHu Laboratory Wuhan China School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China University of Colorado Colorado Springs Department of Computer Science Colorado Springs Colorado United States Institute for Network Sciences and Cyberspace Tsinghua University Beijing China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)9798400702174

Software vulnerabilities are a major cyber threat and it is important to detect them. One important approach to detecting vulnerabilities is to use deep learning while treating a program function as a whole, known as function-level vulnerability detectors. However, the limitation of this approach is not understood. In this paper, we investigate its limitation in detecting one class of vulnerabilities known as inter-procedural vulnerabilities, where the to-be-patched statements and the vulnerability-triggering statements belong to different functions. For this purpose, we create the first Inter -Procedural Vulnerability Dataset (InterPVD) based on C/C++ open-source software, and we propose a tool dubbed VulTrigger for identifying vulnerability-triggering statements across functions. Experimental results show that VulTrigger can effectively identify vulnerability-triggering statements and inter-procedural vulnerabilities. Our findings include: (i) inter-procedural vulnerabilities are prevalent with an average of 2.8 inter-procedural layers;and (ii) function-level vulner-ability detectors are much less effective in detecting to-be-patched functions of inter-procedural vulnerabilities than detecting their counterparts of intra-procedural vulnerabilities. © 2024 ACM.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

A novel coalition game on management of electronic health records in blockchain technology

引用

Journal of Ambient Intelligence and Humanized Computing 2025年第4期16卷 653-666页

作者： Hamidoǧlu, Ali Altiti, Mustafa Jana, Chiranjibe Weber, Gerhard-Wilhelm Department of Mathematical and Statistical Sciences University of Alberta EdmontonABT6G 2J5 Canada Cyber Security Department Bahçeşehir University Çrağan İstanbul34353 Turkey Saveetha School of Engineering Tamil Nadu Chennai602105 India Faculty of Engineering Management Poznan University of Technology Poznan Poland Institute of Applied Mathematics Middle East Technical University Ankara Turkey

The collection and integration of health data through electronic health records (EHRs) has improved the overall quality of healthcare. By centralizing all healthcare institutions’ operations in a single record and making it available to authorized individuals, blockchain applications offer a safe, rapid, and secure way to store EHRs. Healthcare providers require clarity on the specific blockchain technology necessary for managing medical records. In this regard, it is essential to understand this technology in terms of cost, speed, security, block size, and its capacity for handling extensive volumes of EHRs. To cope with this limitation, this study divides blockchain technology into two classes: private blockchain and public blockchain, where each blockchain is characterized in terms of its operation and installation costs, processing speed, block size capacity, and level of security. To provide a rational and robust blockchain environment, doctors and patients involve themselves in a novel coalition game to decide which blockchain technology is beneficial for handling their EHRs. Shapley value analysis introduces two novel concepts: fair blockchain and secure blockchain, providing a rational and reliable decision-making process for doctors and patients to determine their policy toward blockchain application when dealing with their EHRs. The study suggests that under a fair and secure blockchain platform, medical doctors and patients prefer private blockchain to public blockchain when dealing with large-scale EHR data. Furthermore, they exhibit the highest Shapley value within the coalition, indicating a strong motivation to utilize blockchain technology in their EHR interactions. Numerical results are performed based on two sample data sets where theoretical findings are justified. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

An Efficient Key Establishment and Distribution Scheme for Wireless Body Area Networks

An Efficient Key Establishment and Distribution Scheme for W...

引用

Science, engineering and Business for Sustainable Development Goals (SEB-SDG), International Conference on

作者： Yusuf Taofeek Waziri Onomza Victor Morufu Olalere Muhammad Bashir Abdullahi Cyber Security Science Department Federal University of Technology Minna Nigeria Computer Science Department Federal University of Technology Minna Nigeria

ISBN: (数字)9798350358155

ISBN: (纸本)9798350358162

Secure communication in wireless body area network (WBAN) depends on efficient key management system. However, to ensure robustness, key management schemes must not only be lightweight but also support the establishment of communication keys efficiently in this paper, an efficient key management scheme is proposed based on elliptic curve cryptography (ECC) algorithm. The proposed scheme adopts a novel authentication mechanism base on authentication tables We analyzed and compare the resilience and performance of our work with concurrent schemes in terms of computation, storage and communication overheads. The results showed an improvement in our scheme in comparison with related work as displayed in table 4. Therefore, WBAN as an innovative part of the industry 4.0 have revolutionized the health-care industry. The application of the technology has transformed laboratory based medical diagnosis and data collections such as blood pressure, glucose level and blood oxygen through remote monitoring, sensing and transmission to health-care service provider for diagnosing different medical treatments. Thus, the design of such critical system requires secure and efficient technique as proposed in this work.

关键词： Wireless communication Wireless sensor networks Authentication Body area networks Elliptic curve cryptography Robustness Sensors

来源：评论

学校读者我要写书评

暂无评论

Green5G: Enhancing Capacity and Coverage in Device-to-Device Communication

引用

computers, Materials & Continua 2021年第5期67卷 1933-1950页

作者： Abdul Rehman Javed Rabia Abid Bakhtawar Aslam Hafiza Ammara Khalid Mohammad Zubair Khan Omar H.Alhazmi Muhammad Rizwan Department of Cyber Security Air UniversityIslamabadPakistan Department of Computer Science Kinnaird College for WomenLahore54000Pakistan Department of Computer Science College of Computer Science and EngineeringTaibah UniversityMadinahSaudi Arabia

With the popularity of green computing and the huge usage of networks,there is an acute need for expansion of the 5G network.5G is used where energy efficiency is the highest priority,and it can play a pinnacle role in helping every industry to hit *** in the 5G network,conventional performance guides,such as network capacity and coverage are still major issues and need *** to Device communication(D2D)communication technology plays an important role to improve the capacity and coverage of 5G technology using different *** issue of energy utilization in the IoT based system is a significant exploration *** optimizationin D2D communication is an important *** need to resolve this issue for increasing system *** IoT speaks to the issue of lessening energy utilization of IoT gadgets which accomplishes a supportable climate for IoT *** this paper,we improve the capacity and coverage of 5G technology using Multiple Inputs Multiple Outputs(MU-MIMO).MUMIMO increases the capacity of 5G in D2D *** also present all the problems faced by 5G technology and proposed architecture to enhance system performance.

关键词： Green computing 5G energy efficiency resource optimization device to device communication multiple input multiple output

来源：评论

学校读者我要写书评

暂无评论

HSFL: Efficient and Privacy-Preserving Offloading for Split and Federated Learning in IoT Services

HSFL: Efficient and Privacy-Preserving Offloading for Split ...

引用

2023 IEEE International Conference on Web Services, ICWS 2023

作者： Deng, Ruijun Du, Xin Lu, Zhihui Duan, Qiang Huang, Shih-Chia Wu, Jie Fudan University School of Computer Science Shanghai China Ministry of Education Engineering Research Center of Cyber Security Auditing and Monitoring Shanghai China Shanghai Blockchain Engineering Research Center Shanghai China Pennsylvania State University Information Sciences and Technology Department AbingtonPA United States National Taipei University of Technology Department of Electronic Engineering Taiwan

ISBN: (纸本)9798350304855

Distributed machine learning methods like Federated Learning (FL) and Split Learning (SL) meet the growing demands of processing large-scale datasets under privacy restrictions. Recently, FL and SL are combined in hybrid SLFL (SFL) frameworks to exploit both methods' advantages to facilitate ubiquitous intelligence in the Internet of Things (IoT), for example, smart finance. Despite its significant impact on the performance and costs of SFL, model decomposition that splits an ML model into the client-server pair has not been sufficiently studied, especially for SFL in a large-scale dynamic IoT environment. In this paper, we propose a new SFL framework HSFL with a lightweight model decomposition method to offload a part of model training to the edge server. Specifically, we develop a method for estimating the training latency of HSFL and designed a metric for measuring privacy leakage in HSFL, based on which we formulate model decomposition in HSFL as an optimization problem with privacy protection as a constraint. Then, we transform the formulated problem into a contextual bandit problem and design an efficient algorithm to solve it. We have conducted thorough evaluations of the proposed HSFL framework through extensive experiments on a prototype testbed and a simulation platform. The experimental results validate the superiority of HSFL over the state-of-the-art benchmarks in terms of training latency, efficiency, scalability, and privacy protection. © 2023 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Enhancing Convergence Efficiency in Federated Learning Models 25

Enhancing Convergence Efficiency in Federated Learning Model...

引用

25th International Arab Conference on Information technology, ACIT 2024

作者： Samara, Ghassan Elhilo, Ala'Mazen Aljaidi, Mohammad Alamleh, Amneh Al-Mousa, Mohammad Rasmi Batyha, Radwan Alazaidah, Raed Bani-Bakr, Alaa Elrashidi, Ali Zarqa University Faculty of It Department of Computer Science Zarqa Jordan Zarqa University Faculty of Information Technology Zarqa Jordan Zarqa University Faculty of Information Technology Department of Data Science and Ai Zarqa Jordan Zarqa University Department of Cybersecurity Zarqa Jordan Applied Science Private University Faculty of Information Technology Amman Jordan Zarqa University Faculty of It Department of Cyber Security Zarqa Jordan University of Business and Technology Department of Electrical Engineering Jeddah21432 Saudi Arabia

ISBN: (纸本)9798331540012

This study evaluates federated learning models' performance compared to centralized models using the MNIST and CIFAR-10 datasets. Four models were tested: two Federated Learning models and two centralized models. The evaluation focused on accuracy, training rounds, and CO2 emissions, to optimize convergence and reducing energy consumption while maintaining strong model performance. The results showed that Federated learning models outperformed centralized models, with the Federated Learning MNIST model achieving 98.79% accuracy compared to 87.93% for the centralized model. Also, the accuracy of the Federated Learning CIFAR-10 model achieved 91.20% accuracy, exceeding the centralized model's 90.14%. The most accuracy improvements happened in the early training rounds. An adjusted hill-climbing technique was used to the centralized models, decreasing the number of iterations and CO2 emissions without significantly reducing accuracy. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

An Approach for Searching the Top K Nearest Areas Based on R-tree

An Approach for Searching the Top K Nearest Areas Based on R...

引用

2023 International Conference on Networks, Communications and Intelligent Computing, NCIC 2023

作者： Lin, Wenhao Lin, Yifeng Yang, Yuer College of Information Science and Technology Jinan University Guangzhou China College of Cyber Security Jinan University Guangzhou China The University of Hong Kong Department of Computer Science Hong Kong Hong Kong

ISBN: (纸本)9798350371949

The K-nearest neighbor (KNN) fundamental MBR searching is a widely used technique in navigation software. It helps locate regions, such as nearby buildings, that cannot be treated as individual entities. When searching for the top k nearest fundamental MBRs or KNN points in an unevenly distributed space, algorithms based on the R-tree prove to be the most effective ones. R-tree allows for efficient indexing and searching of multidimensional space, making it a powerful tool for spatial querying and top k nearest neighbor searching. In this paper, we compare KNN based on R-tree with linear scanning in 2D graph-based search algorithms. Experimental results demonstrate that our proposed searching algorithms based on R-tree outperform linear scanning by being approximately 95.357% faster in range querying and approximately 97.826% faster in finding the top k nearest fundamental MBRs. Furthermore, the searching algorithms based on the R-tree can be extended to find the top k nearest fundamental MBRs of q in high-dimensional spaces. © 2023 IEEE.

关键词： Nearest neighbor search

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：