检索结果-内蒙古大学图书馆

Deep Learning Approach for Detecting Fake Images Using Texture Variation Network

Inteligencia Artificial 2023年第72期26卷 1-14页

作者： Haseena, S. Saroja, S. Shri Dharshini, D. Nivetha, A. Department of Information Technology Mepco Schlenk Engineering College India Department of Computer Applications National Institute of Technology Trichy India

Face manipulation technology is rapidly evolving, making it impossible for human eyes to recognize fake faces in photos. Convolutional Neural Network (CNN) discriminators, on the other hand, can quickly achieve high accuracy in distinguishing fake from real face photos. In this paper, we investigate how CNN models distinguish between fake and real faces. According to our findings, face forgery detection heavily relies on the variation in the texture of the images. As a result of the aforementioned discovery, we propose a deep texture variation network, a new model for robust face fraud detection based on convolution and pyramid pooling. Convolution combines pixel intensity and pixel gradient information to create a stationary representation of composition difference information. Simultaneously, multi-scale information fusion based on the pyramid pooling can prevent the texture features from being destroyed. The proposed deep texture variation network outperforms previous techniques on a variety of datasets, including Faceforensics++, DeeperForensics-1.0, Celeb-DF, and DFDC. The proposed model is less susceptible to image distortion, such as JPEG compression and blur, which is important in this field. © IBERAMIA and the authors.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

A novel dominating set and centrality based graph convolutional network for node classification

引用

Multimedia Tools and Applications 2025年第14期84卷 12979-13004页

作者： Garg, Neeraj Garg, Sneha Jadeja, Mahipal Meena, Yogesh Kumar Gopalani, Dinesh Chauhan, Ganpat Singh Department of Computer Science and Engineering Malaviya National Institute of Technology Jaipur Rajasthan Jaipur302017 India Department of Computer Science and Engineering Swami Keshvanand Institute of Technology Jaipur Rajasthan Jaipur302017 India Department of Information Technology School of IT Manipal University Jaipur Rajasthan Jaipur302034 India

To learn and analyze graph-structured data, Graph Neural Networks (GNNs) have emerged as a powerful framework over traditional neural networks, which work well on grid-like or sequential structure data. GNNs are particularly designed to record dependencies and relations in large and complicated networks such as social networks, molecular structures, citation networks and recommendation systems. The main idea behind GNN is the message-passing mechanism that allows sharing of information between nodes and edges in a graph network, enabling the model to learn about graph structures and perform well even with limited training data. GNNs have three different architectures: Graph Convolutional Network (GCN), GraphSAGE, Graph Attention Model (GAT) and other hybrid models. In this paper, we have proposed novel GCN models namely dominating Set with Degree Centrality based GCN (DS_DC_GCN) and dominating Set with Betweenness Centrality based GCN (DS_BWC_GCN) model, for node classification task. In the traditional GNN models, training nodes are selected randomly. In our proposed models, we first identify a subset of the nodes with the help of a dominating set (D) of the underlying network. Then we consider final training nodes (say D′) from set D based on the centrality scores (degree or betweenness). We give preference to nodes having high centrality scores for training. To evaluate the performance of the proposed models, we have used three standard citation network datasets - Cora, Citeseer and Pubmed. Our models obtain improved node classification performance as compared to other state-of-the-art-methods. The consideration of high centrality nodes of dominating set in training helps to improve the feature availability in the underlying computational graphs, which in turn improves learning and hence, model performance. These models also have the potential to offer better results for other network tasks, including link prediction and community detection. © The Author(s), under

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Semantic segmentation of brain tumor images using attention-based residual light u-net model

引用

Multimedia Tools and Applications 2025年第10期84卷 7425-7441页

作者： Chakrapani Kumar, Sunil Department of Information Technology ABV-IIITM MP Gwalior474015 India Department of Computer Science and Engineering ABV-IIITM MP Gwalior474015 India

Accurate segmentation of brain tumor regions in MRI images is essential for monitoring tumor growth. In view of this, several automated brain tumor segmentation models are proposed. U-Net is one of the most popular models widely used for image segmentation. Based on the U-Net architecture, complex models are proposed for brain tumor segmentation. However, existing models mainly focus on increasing the depth of the model for better accuracy without much bothering the transmission of spatial information. Attention-based U-Net model can help to extract salient local-level features that can be passed to the decoder part of the network. Considering the above issues, we propose a variant of the U-Net model called the Attention-based Residual Light U-Net model. The proposed model is effective in the sense that it has comparatively less model complexity, a better attention mechanism, and a combined loss function that can handle class imbalance problems. Our experiments on one of the latest BraTS-2021 datasets show that the proposed method can achieve a mean Intersection over Union of 0.8874, 0.8863, 0.8905, and dice score of 0.9415, 0.9387, 0.9115 on the whole tumor, tumor core, and enhancing tumor, respectively. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Tumors

来源：评论

学校读者我要写书评

暂无评论

Towards the Development of a Copyright Risk Checker Tool for Generative Artificial Intelligence Systems

引用

Digital Government: Research and Practice 2024年第4期5卷 1-21页

作者： Billiris, Grace Gill, Asif Oppermann, Ian Niazi, Mahmood Faculty of Engineering and Information Technology University of Technology SydneyNSW Australia Computer Science University of Technology SydneyNSW Australia University of Technology SydneyNSW Australia King Fahd University of Petroleum and Minerals Department of Information and Computer Science Dhahran Saudi Arabia

Generative Artificial Intelligence (GAI) is fundamentally changing the ways of working and blurring the boundaries between human and machine-generated contents. While there is an increasing interest in the adoption of GAI systems, such as ChatGPT and DALL-E, there are also serious concerns about the copyright of the contents - the inputs or generated as outputs by the GAI systems. Such concerns need to be identified and assessed to ensure the ethical and responsible use of GAI systems. Thus, this article aims to address the key research challenge: "how to identify and assess GAI system's copyright concerns"? In response, we propose the development of a Copyright Risk Checker (CRC) Tool. This tool has been formulated and evaluated using a recognised design science research methodology, drawing on an analysis of 10 legal cases across Australia, the United Kingdom, the United States, and Europe. The CRC Tool has undergone evaluation through an experimental scenario, and the results suggest that it is suitable for conducting an indicative copyright risk check of GAI systems. The outcomes of this preliminary assessment can be further examined by expert legal advisors for an in-depth analysis. The development of the CRC Tool provides a foundation for continued research and advancement in this significant area of study. © 2024 Copyright held by the owner/author(s).

关键词： Generative Artificial Intelligence Copyright Concern GAI Governance Copyright Regulations Independent Intellectual Effort Authorship

来源：评论

学校读者我要写书评

暂无评论

Frame aggregation aware cluster based association management for heterogeneous users

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第3期14卷 2811-2826页

作者： Tewari, Babul P. Department of Computer Science and Engineering Indian Institute of Information Technology Bhagalpur813 210 India

In this paper, a cluster based association management for heterogeneous users has been addressed. The aim is to explore frame aggregation for the high throughput (HT) users such as 802.11n type, avoiding the drawback of mixed contention from the presence of the coexisting legacy users such as 802.11 g. It has been demonstrated that restricting an access point (AP) to serve only similar type of users may lead suboptimal performance both in terms of throughput and service coverage as well. Additionally, frame aggregation can not perform well under this common contending scenario. Therefore, a cluster based association management has been proposed which uses frame aggregation and suitably designs a user-AP association where both HT users and legacy users coexist and access services from a common AP. A mathematical formulation of the proposed approach has been developed followed by a greedy solution. Extensive simulations have been performed that establish the fact that the proposed approach judiciously associates the heterogeneous users to the selected target AP and outperforms the other existing approaches including some benchmark algorithms. The results also confirm the effective use of frame aggregation and novelty of the proposed approach in terms of throughput, fairness and delay overhead. © 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Wireless local area networks (WLAN)

来源：评论

学校读者我要写书评

暂无评论

Leveraging Concise Concepts With Probabilistic Modeling for Interpretable Visual Recognition

引用

IEEE Transactions on Multimedia 2025年 27卷 3117-3131页

作者： Zhang, Yixuan Liu, Chuanbin Liu, Yizhi Gao, Yifan Lu, Zhiying Xie, Hongtao Zhang, Yongdong University of Science and Technology of China School of Information Science and Technology Hefei230027 China Hunan University of Science and Technology Department of Computer Science and Engineering Hunan411199 China

Interpretable visual recognition is essential for decision-making in high-stakes situations. Recent advancements have automated the construction of interpretable models by leveraging Visual Language Models (VLMs) and Large Language Models (LLMs) with Concept Bottleneck Models (CBMs), which process a bottleneck layer associated with human-understandable concepts. However, existing methods suffer from two main problems: a) the collected concepts from LLMs could be redundant with task-irrelevant descriptions, resulting in an inferior concept space with potential mismatch. b) VLMs directly map the global deterministic image embeddings with fine-grained concepts results in an ambiguous process with imprecise mapping results. To address the above two issues, we propose a novel solution for CBMs with Concise Concept and Probabilistic Modeling (CCPM) that can achieve superior classification performance via high-quality concepts and precise mapping strategy. First, we leverage in-context examples as category-related clues to guide LLM concept generation process. To mitigate redundancy in the concept space, we propose a Relation-Aware Selection (RAS) module to obtain a concise concept set that is discriminative and relevant based on image-concept and inter-concept relationships. Second, for precise mapping, we employ a Probabilistic Distribution Adapter (PDA) that estimates the inherent ambiguity of the image embeddings of pre-trained VLMs to capture the complex relationships with concepts. Extensive experiments indicate that our model achieves state-of-the-art results with a 6.18% improvement in classification accuracy on eight mainstream recognition benchmarks as well as reliable explainability through interpretable analysis. © 1999-2012 IEEE.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Securing Cloud-Encrypted Data:Detecting Ransomware-as-a-Service(RaaS)Attacks through Deep Learning Ensemble

引用

computers, Materials & Continua 2024年第4期79卷 857-873页

作者： Amardeep Singh Hamad Ali Abosaq Saad Arif Zohaib Mushtaq Muhammad Irfan Ghulam Abbas Arshad Ali Alanoud Al Mazroa School of Computer Information Sciences University of CumberlandsWilliamsburgKentucky40769USA Computer Science Department College of Computer Science and Information SystemsNajran UniversityNajran61441Saudi Arabia Department of Mechanical Engineering College of EngineeringKing Faisal UniversityAl-Ahsa31982Saudi Arabia Department of Electrical Electronics and Computer SystemsCollege of Engineering and TechnologyUniversity of SargodhaSargodha40100Pakistan Electrical Engineering Department College of EngineeringNajran UniversityNajran61441Saudi Arabia Department of Electrical Engineering The University of LahoreLahore54000Pakistan Faculty of Computer and Information Systems Islamic University of MadinahMadinah42351Saudi Arabia Department of Information Systems College of Computer and Information SciencesPrincess Nourah bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia

Data security assurance is crucial due to the increasing prevalence of cloud computing and its widespread use across different industries,especially in light of the growing number of cybersecurity threats.A major and everpresent threat is Ransomware-as-a-Service(RaaS)assaults,which enable even individuals with minimal technical knowledge to conduct ransomware *** study provides a new approach for RaaS attack detection which uses an ensemble of deep learning *** this purpose,the network intrusion detection dataset“UNSWNB15”from the Intelligent Security Group of the University of New South Wales,Australia is *** the initial phase,the rectified linear unit-,scaled exponential linear unit-,and exponential linear unit-based three separate Multi-Layer Perceptron(MLP)models are ***,using the combined predictive power of these three MLPs,the RansoDetect Fusion ensemble model is introduced in the suggested *** proposed ensemble technique outperforms previous studieswith impressive performance metrics results,including 98.79%accuracy and recall,98.85%precision,and 98.80%*** empirical results of this study validate the ensemble model’s ability to improve cybersecurity defenses by showing that it outperforms individual *** expanding the field of cybersecurity strategy,this research highlights the significance of combined deep learning models in strengthening intrusion detection systems against sophisticated cyber threats.

关键词： Cloud encryption RaaS ensemble threat detection deep learning cybersecurity

来源：评论

学校读者我要写书评

暂无评论

Enhancing Ddos Attack Detection and Prevention in Mobile IP Networks Using HCIF 5

Enhancing Ddos Attack Detection and Prevention in Mobile IP ...

引用

5th IEEE Global Conference for Advancement in technology, GCAT 2024

作者： Shirke, Sonali Bharat Jaleel, Uruj Kadam, Sandeep Junnarkar, Aparna Atul Kalinga University Department of Computer Sci. and Engg India Keystone School of Engineering Department of Computer Engineering India Vishwakarma Institute of Information Technology Department of Information Technology India

ISBN: (纸本)9798350376685

This research suggests a Hybrid Cache-based Intelligent Framework (HCIF) to help mobile IP networks better identify and mitigate DDoS attacks. To improve network security and QoS, the HCIF technique integrates source monitoring, attack detection, and verification based on Turing tests. We evaluate HCIF and the state-of-the-art RBPBoost approach in NS-3 simulations, taking into account different mobility situations and attacker numbers. Throughput, packet delivery ratio, and packet loss ratio are all areas where HCIF excels compared to RBPBoost, while delay performance is still an issue. © 2024 IEEE.

关键词： Internet protocols

来源：评论

学校读者我要写书评

暂无评论

Sample-based relationship for assisting diagnosis of pneumonia in medical care

引用

Multimedia Tools and Applications 2025年第1期84卷 267-288页

作者： Chen, Hongkang Lu, Huijuan Zhu, Wenjie Zhou, Ye Yao, Yudong Wang, Renfeng College of Information Engineering China Jiliang University Zhejiang Hangzhou China Drore Information and Technology Co. LTD Zhejiang Hangzhou China Department of Electrical and Computer Engineering Stevens Institute of Technology HobokenNJ United States College of Computer Science and Technology Zhejiang University Zhejiang Hangzhou China

Deep neural networks have succeeded in learning balanced and imbalanced data in the field of pneumonia diagnosis. However, both require separate model designs in their respective domains. The pneumonia recognition methods have poor transfer performance between the two types of image distribution data. This paper proposes BeFormer, which focuses on sample relationships. BeFormer introduces two stages, feature extraction and establishing sample relationships, to solve the problem. Specifically, in the feature extraction stage, a new backbone network called PANet based on Multi-Scale Frequency Domain Attention (FPSA) is proposed. In establishing sample relationships, Deformable Transformer is used in parallel to model the relationships between partial regions of interest among the same and different samples in the pneumonia dataset. These two modules are connected in series to form BeFormer. Without making any changes to the model, this paper conducted experiments using BeFormer on both balanced and imbalanced pneumonia datasets. As compared to traditional pneumonia recognition models, the proposed method achieved significant improvements in the two different data distributions, demonstrating outstanding transfer performance. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Recent Trends in Compact Planar Antennas at 5G Sub-6 GHz and mm Wave Frequency Bands for Automotive Wireless Applications: A Review

引用

Progress In Electromagnetics Research C 2024年 143卷 169-180页

作者： Kumar, Ashish Aljaidi, Mohammad Singh, Manpreet Alshammari, Mohammed Alsuwaylimi, Amjad A. Alenezi, Sami M. Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Department of Computer Science Faculty of Information Technology Zarqa University Zarqa13110 Jordan Department of Computer Sciences Faculty of Computing and Information Technology Northern Border University Rafha91911 Saudi Arabia Department of Information Technology College of Computing and Information Technorlogy Northern Border University Arar91431 Saudi Arabia

5G wireless communication offers higher channel capacity, high data rate, sufficient bandwidth, enhanced coverage, and reliable link as compared to the previous generation mobile networks. Also, 5G becomes more relevant with the recent launches of low earth orbit satellites by various ventures like starlink, amazon, one web, etc. As these satellites have heights up to the range from 300 km to 1000 km, the free space path loss decreases drastically which in turn improves the signal reliability and efficient communication at dead spots. With these advancements, antenna researchers have the freedom to design compact, easy to manufacture, with adequate gain user equipment terminal antennas which can be easily integrated in the modern passenger car body. This paper will focus on the recent development in the design considerations of the compact antenna design at sub-6 GHz (n78 frequency band) and mm wave (n278 frequency band) according to 3gpp standards. This manuscript also discusses various design specifications like selection of material for antennas, design complexity, feeding methods, fabrication and measurement challenges and performance parameters which include reflection coefficient, gain, polarization, axial ratio, cross polarization discrimination, radiation efficiency, radiation pattern shape, etc. The proposed antenna design considerations will facilitate the possible integration into the various parts of the car body according to the recent vehicular applications for uninterrupted communication. © 2024, Electromagnetics Academy. All rights reserved.

关键词： Radiation efficiency

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：