检索结果-内蒙古大学图书馆

Edge Graph Intelligence: Reciprocally Empowering Edge Networks With Graph Intelligence

学校读者我要写书评

暂无评论

IEEE Communications Surveys and Tutorials 2025年

作者： Zeng, Liekang Ye, Shengyuan Chen, Xu Zhang, Xiaoxi Ren, Ju Tang, Jian Yang, Yang Shen, Xuemin Sherman Sun Yat-sen University School of Computer Science and Engineering Guangzhou510275 China Guangzhou511453 China Ministry of Education School of Computer Science and Engineering Sun Yat-sen University The Key Laboratory of Machine Intelligence and Advanced Computing Guangzhou510275 China Tsinghua University Department of Computer Science and Technology Beijing100084 China Midea Group Shanghai201702 China Guangzhou511453 China Peng Cheng Laboratory Shenzhen518055 China Terminus Group Beijing100027 China University of Waterloo Department of Electrical and Computer Engineering WaterlooONN2L 3G1 Canada

Recent years have witnessed a thriving growth of computing facilities connected at the network edge, cultivating edge networks as a fundamental infrastructure for supporting miscellaneous intelligent services. Meanwhile, Artificial Intelligence (AI) frontiers have extrapolated to the graph domain and promoted Graph Intelligence (GI). Given the inherent relation between graphs and networks, the interdiscipline of graph learning and edge networks, i.e., Edge GI or EGI, has revealed a novel interplay between them - GI aids in optimizing edge networks, while edge networks facilitate GI model deployment. Driven by this delicate closed-loop, EGI is recognized as a promising solution to fully unleash the potential of edge computing power and is garnering growing attention. Nevertheless, research on EGI remains nascent, and there is a soaring demand within both the communications and AI communities for a dedicated venue to share recent advancements. To this end, this paper promotes the concept of EGI, explores its scope and core principles, and conducts a comprehensive survey concerning recent research efforts on this emerging field. Specifically, this paper introduces and discusses: 1) fundamentals of edge computing and graph learning, 2) emerging techniques centering on the closed loop between graph intelligence and edge networks, and 3) open challenges and research opportunities of future EGI. By bridging the gap across communication, networking, and graph learning areas, we believe that this survey can garner increased attention, foster meaningful discussions, and inspire further research ideas in EGI. © 1998-2012 IEEE.

关键词： Contrastive Learning

An Accurate Iris Recognition System Based on Deep Transfer Learning and Data Augmentation

学校读者我要写书评

暂无评论

An Accurate Iris Recognition System Based on Deep Transfer L...

Innovative and Intelligent Information Technologies (IC3IT), International Conference on

作者： Abdessalam Hattab Ali Behloul Wassila Hattab Computer Science Department LaSTIC Laboratory Batna 2 university Batna Algeria Laboratory of Biskra LGEB Electricals Engineering Biskra University Biskra Algeria

ISBN: (数字)9798331541064

ISBN: (纸本)9798331541071

Iris recognition is a highly secure biometric identification method that utilizes the distinct and intricate patterns of the iris, which remain consistent throughout a person's lifetime. Remarkably, it can even distinguish between identical twins, highlighting its precision and reliability. However, despite these advantages, challenges such as reflections, occlusions, varying lighting conditions, blur, and other environmental factors can significantly degrade the performance of iris recognition systems. Deep Convolutional Neural Networks (CNNs), known for their success in image recognition, often require numerous parameters, increasing computational time and resource demands. To address these challenges, we propose an advanced iris recognition system based on Deep CNNs. Our system utilizes YOLOv4-tiny for accurate iris region detection, and for the recognition task, we introduce a deep CNNs model inspired by the pre-trained Xception, which contains fewer than three million parameters. To effectively train this deep CNN on small iris datasets, we apply Data Augmentation and Transfer Learning techniques, which improve accuracy while reducing the risk of overfitting. Through two-fold cross-validation, our system achieves impressive accuracy rates of 99.81 % and 99.64 % on the CASIA-Iris-Interval and IITD datasets, respectively. Our method demonstrates a clear advantage over existing methods in terms of accuracy and reliability, making it a promising solution for various applications such as border control, law enforcement, and access control.

关键词： Accuracy Face recognition Transfer learning Lighting Data augmentation Reflection Data models Reliability Iris recognition Overfitting

LENS-XAI: Redefining Lightweight and Explainable Network Security through Knowledge Distillation and Variational Autoencoders for Scalable Intrusion Detection in Cybersecurity

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Yagiz, Muhammet Anil Goktas, Polat Department of Computer Engineering Kırıkkale University Kırıkkale71450 Turkey School of Computer Science & CeADAR University College Dublin DublinD04 V2N9 Ireland

The rapid proliferation of Industrial Internet of Things (IIoT) systems necessitates advanced, interpretable, and scalable intrusion detection systems (IDS) to combat emerging cyber threats. Traditional IDS face challenges such as high computational demands, limited explainability, and inflexibility against evolving attack patterns. To address these limitations, this study introduces the Lightweight Explainable Network Security framework (LENS-XAI), which combines robust intrusion detection with enhanced interpretability and scalability. LENS-XAI integrates knowledge distillation, variational autoencoder models, and attribution-based explainability techniques to achieve high detection accuracy and transparency in decision-making. By leveraging a training set comprising 10% of the available data, the framework optimizes computational efficiency without sacrificing performance. Experimental evaluation on four benchmark datasets-Edge-IIoTset, UKM-IDS20, CTU-13, and NSL-KDD-demonstrates the framework's superior performance, achieving detection accuracies of 95.34%, 99.92%, 98.42%, and 99.34%, respectively. Additionally, the framework excels in reducing false positives and adapting to complex attack scenarios, outperforming existing state-of-the-art methods. Key strengths of LENS-XAI include its lightweight design, suitable for resource-constrained environments, and its scalability across diverse IIoT and cybersecurity contexts. Moreover, the explainability module enhances trust and transparency, critical for practical deployment in dynamic and sensitive applications. This research contributes significantly to advancing IDS by addressing computational efficiency, feature interpretability, and real-world applicability. Future work could focus on extending the framework to ensemble AI systems for distributed environments, further enhancing its robustness and adaptability. © 2025, CC BY.

关键词： Industrial internet of things (IIoT)

Surveillance Video Key Frame Extraction Based on Center Offset

学校读者我要写书评

暂无评论

computers, Materials & Continua 2021年第9期68卷 4175-4190页

作者： Yunzuo Zhang Shasha Zhang Yi Li Jiayu Zhang Zhaoquan Cai Shui Lam School of Information Science and Technology Shijiazhuang Tiedao UniversityShijiazhuang050043 China 2Department of Computer Science and Engineering Huizhou UniversityHuizhou516007 China 3College of Engineering California State UniversityLong BeachCA90185USA

With the explosive growth of surveillance video data,browsing videos quickly and effectively has become an urgent *** key frame extraction has received widespread attention as an effective ***,accurately capturing the local motion state changes of moving objects in the video is still challenging in key frame *** target center offset can reflect the change of its motion *** observation proposed a novel key frame extraction method based on moving objects center offset in this *** proposed method utilizes the center offset to obtain the global and local motion state information of moving objects,and meanwhile,selects the video frame where the center offset curve changes suddenly as the key *** processing effectively overcomes the inaccuracy of traditional key frame extraction ***,extracting the center point of each ***,calculating the center point offset of each frame and forming the center offset curve by connecting the center offset of each ***,extracting candidate key frames and optimizing them to generate final key *** experimental results demonstrate that the proposed method outperforms contrast methods to capturing the local motion state changes of moving objects.

关键词： Center offset local motion key frame extraction moving object detection

Dual-pronged deep learning preprocessing on heterogeneous platforms with CPU, GPU and CSD

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wei, Jia Zhang, Xingjun Pedrycz, Witold Wang, Longxiang Zhao, Jie The School of Computer Science and Technology Xi'an Jiaotong University Xi'An710049 China The Department of Electrical and Computer Engineering University of Alberta EdmontonABT6G 2G6 Canada The Department of Electrical and Computer Engineering King Abdulaziz University Jeddah21589 Saudi Arabia The Systems Research Institute PAS Warsaw01-447 Poland The College of Computer Science and Electronic Engineering Hunan University Changsha410082 China

As accelerator computation speeds increase and the number of single compute node accelerators increases, data reading and preprocessing gradually become a bottleneck for deep learning. Most existing data preprocessing is done at the CPU. Although some studies use techniques such as multiprocessing and double buffering to accelerate CPU preprocessing, CPU computational speed and storage bandwidth still limit the processing speed. Other studies try to use intelligent data storage devices, such as computational storage devices (CSD), to complete data preprocessing instead of CPUs. The current studies use only one device to complete data preprocessing operations, which cannot fully overlap data preprocessing and accelerator computation time. To fully exploit the independence and high bandwidth of the novel CSD, this paper proposes an advanced, highly parallel dual-pronged data preprocessing algorithm (DDLP) that significantly improves the execution efficiency and computational overlap between heterogeneous devices. DDLP enables the CPU and CSD to start data preprocessing operations from both ends of the dataset separately. Meanwhile, we propose two adaptive dynamic selection strategies to make DDLP control the GPU to automatically read data from different sources. We achieve sufficient computational overlap between CSD data preprocessing and CPU preprocessing, GPU computation, and GPU data reading. In addition, DDLP leverages GPU Direct Storage (GDS) technology to enable efficient SSD-to-GPU data transfer. DDLP reduces the usage of expensive CPU and DRAM resources, reduces the number of SSD-to-GPU data transfers, and improves the energy efficiency of preprocessing while reducing the overall preprocessing and training time. Extensive experimental results show that DDLP can improve learning speed by up to 23.5% on ImageNet Dataset while reducing energy consumption by 19.7% and CPU and DRAM usage by 37.6%. DDLP also improve learning speed by up to 27.6% on Cifar-10 Dataset

关键词： Dynamic random access storage

Network Traffic Co-Movement Assessment via Oriented Basis Signal Processing and Ensemble Decision Trees 27

学校读者我要写书评

暂无评论

Network Traffic Co-Movement Assessment via Oriented Basis Si...

27th IEEE International Conference on Intelligent Transportation Systems, ITSC 2024

作者： Mohebbi, Mohammad Reza Hassan Zada, Mohammad Javad Rostami-Shahrbabaki, Majid Doller, Mario Yamnenko, Iuliia University of Applied Sciences Josef Ressel Center Vision2Move Project Tirol Kufstein Austria University of Passau Department of Computer Science Germany Technical University of Munich Traffic Engineering and Control Germany National Technical University of Ukraine Igor Sikorsky Kyiv Polytechnic Institute Department of Electronic Devices and Systems Kyiv Ukraine

ISBN: (纸本)9798331505929

Accurate short-term network-wide traffic prediction is essential to guarantee high service quality in urban traffic control systems. Nevertheless, traffic state time series represent network-scale spatiotemporal co-movement patterns and location-specific features. Therefore, hybrid statistical machine learning (ml) algorithms could be utilized to accommodate the aforementioned characteristics. In this paper, a hybrid random forest (rf) and extreme gradient boosting tree (XGBoost) model is introduced for network-wide traffic prediction. Moreover, a multi-noise oriented basis wavelet transform (OBT) filter is employed to pre-process the original time series, and improve the predictive accuracy. The RF model captures the traffic co-movements and the XGBoost extracts the local information. Comparative analysis of the hybrid algorithm and deep learning-based benchmarks indicate better performance of the proposed methodology in hourly traffic state prediction of 30 loop detectors, located in the paris city center. Hence, the hybrid decision-tree-based mechanism is a useful framework for real-time network traffic forecasting offering fewer trainable (hyper-)parameters, and thus, lower computational cost. © 2024 IEEE.

关键词： Time series

ViewFusion: Learning Composable Diffusion Models for Novel View Synthesis

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Spiegl, Bernard Perin, Andrea Deny, Stéphane Ilin, Alexander Department of Computer Science Aalto University Finland Department of Neuroscience and Biomedical Engineering Aalto University Finland System 2 AI Finnish Center for Artificial Intelligence FCAI

Deep learning is providing a wealth of new approaches to the old problem of novel view synthesis, from Neural Radiance Field (NeRF) based approaches to end-to-end style architectures. Each approach offers specific strengths but also comes with specific limitations in their applicability. This work introduces ViewFusion, a state-of-the-art end-to-end generative approach to novel view synthesis with unparalleled flexibility. ViewFusion consists in simultaneously applying a diffusion denoising step to any number of input views of a scene, then combining the noise gradients obtained for each view with an (inferred) pixel-weighting mask, ensuring that for each region of the target scene only the most informative input views are taken into account. Our approach resolves several limitations of previous approaches by (1) being trainable and generalizing across multiple scenes and object classes, (2) adaptively taking in a variable number of pose-free views at both train and test time, (3) generating plausible views even in severely undetermined conditions (thanks to its generative nature)—all while generating views of quality on par or even better than state-of-the-art methods. Limitations include not generating a 3D embedding of the scene, resulting in a relatively slow inference speed, and our method only being tested on the relatively small dataset NMR. Code is available. © 2024, CC BY.

关键词： Deep learning

A Review on Explainable Machine Learning-Driven Systems to Detect Parkinson's Disease from Speech Signals

学校读者我要写书评

暂无评论

SSRN

SSRN 2025年

作者： Manaog, Marianne Lyne Parisi, Luca University of Essex School of Computer Science and Electronic Engineering Wivenhoe Park ColchesterCO4 3SQ United Kingdom Department of Computer Science 5 South Charlotte Street EdinburghEH2 4AN United Kingdom

Parkinson’s Disease (PD) is a neurodegenerative condition that impairs motor function, mental health, and speech, affecting over 8.5 million people globally. Its prevalence has doubled in the past 25 years, and there is currently no cure. Effective management relies on early detection, as treatments like pharmacological interventions and rehabilitation are most beneficial at early stages. However, traditional clinical assessments and AI-based detection methods often diagnose PD at later stages, limiting treatment success. To address this, explainable AI-driven models, such as the Gradient Boosting Trees (GBT) classification system, are recommended to be leveraged on speech data. Storing them on an Apache Cassandra database would enable to identify patterns as they evolve over time more proactively than when in other databases, being able to detect them from their longitudinal view. Data on speech signals and metrics must meet the minimum sample size-related requirements as determined via a statistical power analysis, ensuring that the AI-driven systems’ results are evaluated reliably. Classification models’ predictive performance should be evaluated at least based on accuracy, and other metrics that consider class imbalances that are typical of PD-related datasets, such as the weighted precision, the weighted recall, and the weighted F1-score on the test set for assessing generalisation. Furthermore, vector similarity was identified as being a catalyst to supports real-time PD detection performed via AI-driven systems. AI-driven systems need to focus on improving early PD detection and management, thus offering a proactive solution to a significant healthcare challenge. © 2025, The Authors. All rights reserved.

关键词： Neurodegenerative diseases

Determining Disturbance Recovery Conditions by Inverse Sensitivity Minimization

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Fisher, Michael W. Hiskens, Ian A. Department of Electrical and Computer Engineering University of Waterloo 200 University Avenue West WaterlooONN2L 3G1 Canada Department of Electrical Engineering and Computer Science University of Michigan 1301 Beal Avenue Ann ArborMI48109 United States

Power systems naturally experience disturbances, some of which can damage equipment and disrupt consumers. It is important to quickly assess the likely consequences of credible disturbances and take preventive action, if necessary. However, assessing the impact of potential disturbances is challenging because many of the influential factors, such as loading patterns, controller settings and load dynamics, are not precisely known. To address this issue, the paper introduces the concept of parameter-space recovery regions. For each disturbance, the corresponding recovery region is the region of parameter space for which the system will recover to the desired operating point. The boundary of the recovery region establishes the separation between parameter values that result in trouble-free recovery and those that incur undesirable non-recovery. The safety margin for a given set of parameter values is defined as the smallest distance (in parameter space) between the given values and the recovery boundary. Novel numerical algorithms with theoretical guarantees are presented for efficiently computing recovery boundaries and safety margins. Unlike prior methods, which tend to be overly conservative and restricted to low dimensional parameter space, these methods compute safety margins to arbitrary user-specified accuracy and do so efficiently in high dimensional parameter space. The efficacy of the methods is demonstrated using the IEEE 39-bus benchmark power system, where safety margins are computed for cases that consider up to 86 parameters, and reveal unexpected safety implications that would not have been observed *** Codes 93D09, 93D99, 37N99 © 2025, CC BY.

关键词： Inverse problems