检索结果-内蒙古大学图书馆

MF-CGAN: Multifeature Conditional GAN for Synthetic Data Generation in Internet of Medical Things

IEEE Internet of Things Journal 2025年第10期12卷 13469-13476页

作者： Pandey, Chandrasen Tiwari, Vaibhav Francis, Sharmila A. J. Pal, Vipin Sinha Roy, Diptendu Manipal University Jaipur Department of Computer Science and Engineering Jaipur303007 India Bennett University School of Computer Science Engineering and Technology Greater Noida201310 India King Khalid University Department of Computer Science and Engineering Abha61421 Saudi Arabia Netaji Subhash University of Technology Department of Computer Science and Engineering New Delhi110078 India National Institute of Technology Meghalaya Department of Computer Science and Engineering Sohra 793108 India

Synthetic data generation via generative artificial intelligence (GenAI) is essential for enhancing cybersecurity and safeguarding privacy in the Internet of Medical Things (IoMT) and healthcare. We introduce multifeature conditional generative adversarial network (GAN) (MF-CGAN), a Conditional GAN with multifeature integration, that generates realistic synthetic data using the WUSTL EHMS 2020 dataset of network traffic and health metrics. MF-CGAN’s architecture integrates conditional variables such as cyberattack types, traffic direction, and status flags, preserving their intricate interdependencies. The generator employs dense layers with batch normalization and Leaky ReLU activations to produce conditional synthetic data;the discriminator evaluates authenticity using the same conditional inputs. Experimental results show that MF-CGAN produces high-fidelity synthetic data, closely mirroring real data characteristics. In addition, we quantitatively demonstrate (MAE = 0.012 , RMSE = 0.035 , and classification accuracy up to 99%) that MF-CGAN-generated data effectively supports cybersecurity tasks, indicating its potential to significantly enhance data availability for IoMT. © 2014 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Halo reduction multi-exposure image fusion technique

引用

Multimedia Tools and Applications 2025年第13期84卷 12347-12370页

作者： Sharif, Rizwan Amin, Benish Sukhia, Komal Nain Department of Electrical Engineering Institute of Space Technology Islamabad Pakistan Department of Computer Science Institute of Space Technology Islamabad Pakistan

Multi-exposure image fusion (MEF) involves combining images captured at different exposure levels to create a single, well-exposed fused image. MEF has a wide range of applications, including low light, low contrast, night photography, medical imaging, and remote sensing. However, MEF methods often face issues like artifacts, halos around edges, color inconsistencies, noise amplification, and difficulty in preserving fine details. Moreover, assessing the quality of fused images objectively is complex due to the subjective nature of human perception. Solving the challenges is essential to developing efficient MEF techniques that produce high-quality results across various scenarios. The proposed technique introduces an approach to handling halo artifacts and implementing MEF. The Dense Scale-Invariant Feature Transform (DSIFT) is used to capture vital information about image brightness, texture, and edges from source images. Three weight maps are computed from the local mean, signal strength, and the global gradient for initial weight estimation. The local mean represents the brightness of specific image areas, signal strength preserves essential details like textures and edges while reducing image noise, and global gradient helps identify regions with significant pixel value shifts across multiple exposure images. The weight maps are then combined using a weighted average and a Gaussian smoothing filter is applied to reduce inherent noise and discontinuities in the original weights. Subsequently, pyramid decomposition is performed to generate a fused image. The efficiency of the proposed method is extensively tested on challenging multi-exposure image sequences. The results of the proposed approach demonstrate its superiority in both subjective evaluation and objective metrics like the MEF-Structural Similarity Index (MEF-SSIM), Natural Image Quality Evaluator (NIQE) and Gradient based performance measure (QAB/f). © The Author(s), under exclusive licence to Springer

关键词： Image fusion

来源：评论

学校读者我要写书评

暂无评论

CNN-Transformer-Based Cooperative Spectrum Sensing in Cognitive Radio Networks

引用

IEEE Wireless Communications Letters 2025年第5期14卷 1576-1580页

作者： Fang, Xiangqian Jin, Ming Guo, Qinghua Jiang, Tao Ningbo University Faculty of Electrical Engineering and Computer Science Ningbo315211 China University of Wollongong School of Electrical Computer and Telecommunications Engineering WollongongNSW2522 Australia

Recently, deep learning (DL)-based data-driven spectrum sensing has attracted much attention due to its powerful capabilities to exploit underlying signal features. Existing DL-based cooperative spectrum sensing (CSS) detectors usually extract the signal features from each secondary user (SU) individually, followed by feature fusion at a central fusion center (FC). However, this cannot extract correlation features across SUs. To address this limitation, this letter proposes a CNN-Transformer neural network to exploit the correlations among the received signals across and within SUs. Leveraging the CNN-Transformer, we propose a posteriori probability ratio detector, which is optimal for the binary hypothesis testing problem under the Neyman-Pearson criterion. Moreover, unlike most DL-based CSS detectors, the proposed detector considers imperfect and independent reporting channels between SUs and the FC, and exploits the temporal correlation of signals among neighboring samples. Numerical results are provided to demonstrate the superior performance of the proposed detector, showing an improvement of approximately 6 dB over traditional model-driven energy detectors. Moreover, the proposed detector improves the detection probability to 0.99 from 0.88 of data-driven detectors under Rician fading channels. © 2012 IEEE.

关键词： Fading channels

来源：评论

学校读者我要写书评

暂无评论

A Dynamic Characteristic Aware Index Structure Optimized for Real-world Datasets

引用

ACM Transactions on Storage 2025年第2期21卷 1-32页

作者： Yang, Jin Yoon, Heejin Yun, Gyeongchan Noh, Sam H. Choi, Young-Ri Computer Science and Engineering Unist Ulsan Korea Republic of Department of Computer Science Virginia Tech Blacksburg United States

Many datasets in real life are complex and dynamic, that is, their key densities are varied over the whole key space and their key distributions change over time. It is challenging for an index structure to efficiently support all key operations for data management, in particular, search, insert, and scan, for such dynamic datasets. In this article, we present DyTIS (Dynamic dataset Targeted Index Structure), an index that targets dynamic datasets. DyTIS, although based on the structure of Extendible hashing, leverages the CDF of the key distribution of a dataset, and learns and adjusts its structure as the dataset grows. The key novelty behind DyTIS is to group keys by the natural key order and maintain keys in sorted order in each bucket to support scan operations within a hash index. We also define what we refer to as a dynamic dataset and propose a means to quantify its dynamic characteristics. Our experimental results show that DyTIS provides higher performance than the state-of-the-art learned index for the dynamic datasets considered. We also analyze the effects of the dynamic characteristics of datasets, including sequential datasets, as well as the effect of multiple threads on the performance of the indexes. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Dynamic datasets index structure key distribution

来源：评论

学校读者我要写书评

暂无评论

A community-based simulated annealing approach with a new structure-based neighborhood search to identify influential nodes in social networks

引用

Soft Computing 2025年第3期29卷 1567-1585页

作者： Abyaneh, Farzaneh Rajaee Charkari, Nasrollah Moghadam Roayaei, Mehdy Department of Electrical and Computer Engineering Tarbiat Modares University Tehran Iran Faculty of Electrical and Computer Engineering Tarbiat Modares University Shahid Chamran Highway Tehran Iran

Identifying influential nodes has attracted the attention of many researchers in recent years. Because of the weak tradeoff between accuracy and running time, and ignoring the community structure by the proposed algorithms in the past research studies, further studies in this area are required. In this paper, we consider communities and also use a novel structure-based neighborhood search to improve exploration strategy of the simulated annealing (SA) algorithm. Moreover, we use the k-shell method for generating a better initial solution instead of random generation. In the proposed algorithm called Ckshell-SA, first, the communities are detected, then the k-shell method is used in each community to find initial candidate nodes locally. Finally, SA algorithm is applied with a neighborhood search that considers the structural properties of the network, and three centralities to find the influential nodes globally. A derivative of the Ckshell-SA method called kshell-SA is also introduced in this paper to examine the impact of considering communities. Unlike the Ckshell-SA, the community structure is neglected, and the k-shell is performed on the whole network in kshell-SA algorithm. Extensive experiments are conducted on eight real-world networks under Independent Cascade Model (IC) and Weighted Independent Cascade Model (WC). The results show that the Ckshell-SA and kshell-SA algorithms outperform the state-of-the-art algorithms concerning influence spread. Furthermore, the results show that Ckshell-SA is more efficient in networks like Facebook with a high Power Law exponent and higher modularity. On the contrary, kshell-SA is more successful in networks like Slashdot or Epinions with lower modularity. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Simulated annealing

来源：评论

学校读者我要写书评

暂无评论

Multi-Modal Self-Supervised Learning AlgorithmBased Product Recommendation

引用

IEEE Transactions on Automation science and engineering 2025年 22卷 15371-15380页

作者： Gao, Li Fan, Haiping Chen, Qingkui Yang, Heyu School of Optical-Electrical and Computer Engineering University of Shanghai for Science and Technology Shanghai China

With the rise of e-commerce, personalized recommendation algorithms have received much attention in recent years. Meanwhile, multimodal recommendation algorithms have become the next competitive track of personalized recommendation. The current mainstream recommendation algorithms (e.g., NeuMF, NGCF) rely too much on empirical data and do not effectively utilize multimodal information to model user preferences, while there is the problem of sparsity of user preference data. Secondly, they are still limited to using the same weights to fuse the item features of different modalities, and have the problem of inaccurate recommendation. For this reason, this paper proposes Multi-Modal Self-Supervised Learning Algorithm Based Product Recommendation (MMSLRec). (1) Data augmentation by self-supervised adversarial perturbation generation learning is performed to supplement the missing interaction information of tailed products to solve the common data sparsity problem in product recommendation. (2) A multimodal representation learning module is introduced to capture deep features in image and text modalities using AlexNet and Bert, respectively, in order to avoid the recommendation accuracy degradation problem caused by insufficient data feature learning. (3) Learning the user's perceived strength in different modalities, using the user co-occurrence matrix to fuse the features and make recommendations. Experiments on Amazon-Electronics, Amazon-Clothing, and Amazon-Baby, with different degrees of data sparsity verify the effectiveness of our MMSLRec in solving the data sparsity problem and the multimodal user preference modeling problem, and compared to the state-of-the-art baseline recommendation model on the three datasets, the Recall@20 improved by 9.26%, 3.11% and 2.92%, respectively. © 2004-2012 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Energy- and Area-Efficient CMOS Neuron and Max Pooling Circuit for RRAM-Based CNN Accelerators

引用

IEEE Access 2025年 13卷 84329-84340页

作者： Kim, Dowon Lee, Byung-Geun Gwangju Institute of Science and Technology School of Electrical Engineering and Computer Science Gwangju61005 Korea Republic of

This study proposes an energy- and area-efficient CMOS neuron and analog max pooling circuit for RRAM-based convolutional neural network (CNN) accelerators. The proposed max pooling circuit implements a 2 × 2 max pooling operation with a simple analog circuit in the current domain. The currentmode max pooling is performed before the current-to-voltage and voltage-to-digital conversions, thus reducing the number of circuits required for the conversions in the conventional CNN accelerators. Consequently, the 2 × 2 pooling circuit reduces the energy and area of the accelerator to a quarter. In addition to the max pooling circuit, the neuron circuit improves energy efficiency. The proposed neuron circuit implements the ReLU activation function along with digital conversion by modifying the operating principle of a conventional single-slope analog-to-digital converter. By implementing the ReLU function and digital conversion, the neuron achieves high energy efficiency by applying a high-speed asynchronous operation. A prototype chip, which included the neuron and current-mode max pooling circuit, was fabricated using a 130 nm standard CMOS process. The measurement results of the chip demonstrate its energy efficiency of 5.5 TOPS/W. Additionally, a simulation for image recognition demonstrates that a CNN accelerator employing the proposed neuron and max pooling circuit achieves 96.2 and 82.7% classification accuracies for the MNIST and Fashion-MNIST datasets, respectively. © 2013 IEEE.

关键词： RRAM

来源：评论

学校读者我要写书评

暂无评论

FIAC: fine-grained access control mechanism for cloud based IoT framework

引用

International Journal of Grid and Utility Computing 2025年第3期16卷 269-278页

作者： Chaudhury, Bhagwat Prasad Dhal, Kasturi Pattnaik, Srikanta Nayak, Ajit Kumar Department of Computer Science and Engineering Silicon Institute of Technology Odisha Bhubaneswar India Department of Computer Science and Engineering S‘O’A University Odisha Bhubaneswar India

Cloud computing technology provides various computing resources on demand to users on pay per use basis. The technology fails in terms of its usage due to confidentiality and privacy issues. Access control mechanisms are the tools to prevent unauthorized access to remotely stored data. Cipher Text Policy Attribute-based Encryption (CPABE) is a widely used tool for facilitating authorised users to access the remotely stored encrypted data with fine-grained access control. In the proposed model FIAC, access control mechanism is embedded in cloud-based application. The major contribution of this work is the design of three algorithms all of which are attribute based: Key Generation algorithm, and Encryption and Decryption algorithms. Only authorised users can view them to take appropriate action plans. To enhance security of the cloud-based monitoring system, we have embedded a security scheme, all of which are attribute based. Further, the computation time of the model is found to be encouraging so that it can be used in low power devices. The experimental outcomes establish the usability of our model. Copyright © 2025 Inderscience Enterprises Ltd.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

ElHBiAt: Electra pre-training network Hybrid of BiLSTM, the Attention layer to Aspect-Based Sentiment Analysis

引用

IEEE Access 2025年 13卷 88342-88370页

作者： Ghanee, Amin Jahanbin, Kia Yamchi, Abolfazl Razzaghi Shahid Beheshti University Department of Computer Engineering Tehran Iran Yazd University Department of Computer Engineering Yazd Iran Shiraz University Department of Computer Science Engineering and Information Technology Shiraz Iran

Aspect-based sentiment analysis is one of the famous and practical subjects in natural language processing. Traditional sentiment analysis assigns a polarity to the whole text or document and does not consider the aspects of the text, so it cannot correctly show the author’s opinion. Extracting aspects and determining their polarity has many challenges, such as the lack of rich datasets, negative transfer in cross-domain datasets, and fine-tuning of deep models of the ratio of texts with different contexts. In this research, a model has been proposed for aspect-based sentiment analysis with different contexts. This model uses the Electra as the backbone of the architecture, which is trained with different cross-domain datasets to transfer knowledge to the target domain. Also, in the proposed model of ElHBiAt, the resampling method is used to deal with imbalanced datasets, which causes the neural model to find less orientation than the polarities in the majority. In this model, two layers of attention are used to extract the essential aspects of the text, and the deep neural network BiLSMT is used to process the text in a two-way. The results show that the ElHBiAt model has been able to improve the efficiency by an average of 5.78% in the SemEval dataset, by 2% in the SEntFiN dataset, and by an average of 4.3% in the Drug dataset in terms of the F1-score. © 2013 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

Poor and rich squirrel algorithm-based Deep Maxout network for credit card fraud detection

引用

International Journal of Wireless and Mobile Computing 2025年第2期28卷 123-134页

作者： Paul, Annu Paul, Varghese Department of Computer Science Alphonsa College Kottayam Kerala Pala India Department of Computer Science and Engineering Rajagiri School of Engineering and Technology Kerala Kochi India

This paper proposes a Poor and Rich Squirrel Algorithm (PRSA)-based Deep Maxout network to find fraud data transactions in the credit card system. Initially, input transaction data is passed to the data transformation phase, transforming data using Yeo-Johnson (YJ) transformation. Then, the feature selection procedure is done by the Fisher score for creating the unique and significant features. Next, based on the selected textures, the data augmentation mechanism is done using the oversampling model. At last, the fraud detection is carried out by the Deep Maxout network, which is trained by the proposed PRSA optimisation algorithm, derived by integrating Poor and Rich Optimisation (PRO) and Squirrel Search Algorithm (SSA). The integration of parametric features of the PRSA algorithm effectively trained the classifier to update weights to generate the best solution by considering fitness measures. The proposed method achieved the best accuracy, sensitivity, and specificity measures of 0.96, 0.95 and 0.94, respectively. Copyright © 2025 Inderscience Enterprises Ltd.

关键词： credit card data augmentation data transformation deep learning fraud detection

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：