检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Yao, Dezhong Shi, Yuexin Liu, Tongtong Xu, Zhiqiang National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Mohamed bin Zayed University of Artificial Intelligence United Arab Emirates

Federated Learning (FL) is increasingly adopted in edge computing scenarios, where a large number of heterogeneous clients operate under constrained or sufficient resources. The iterative training process in conventional FL introduces significant computation and communication overhead, which is unfriendly for resource-constrained edge devices. One-shot FL has emerged as a promising approach to mitigate communication overhead, and model-heterogeneous FL solves the problem of diverse computing resources across clients. However, existing methods face challenges in effectively managing model-heterogeneous one-shot FL, often leading to unsatisfactory global model performance or reliance on auxiliary datasets. To address these challenges, we propose a novel FL framework named FedMHO, which leverages deep classification models on resource-sufficient clients and lightweight generative models on resource-constrained devices. On the server side, FedMHO involves a two-stage process that includes data generation and knowledge fusion. Furthermore, we introduce FedMHO-MD and FedMHO-SD to mitigate the knowledge-forgetting problem during the knowledge fusion stage, and an unsupervised data optimization solution to improve the quality of synthetic samples. Comprehensive experiments demonstrate the effectiveness of our methods, as they outperform state-of-the-art baselines in various experimental setups. Our code is available at https://***/YXShi2000/FedMHO. © 2025, CC BY.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Inductively coupled Josephson junctions: a platform for rich neuromorphic dynamics

arXiv

引用

arXiv 2025年

作者： Baxevanis, G. Hizanidis, J. School of Electrical and Computer Engineering Aristotle University of Thessaloniki Thessaloniki54124 Greece Institute of Electronic Structure and Laser Foundation for Research and Technology-Hellas Heraklion70013 Greece Institute of Nanoscience and Nanotechnology National Center for Scientific Research "Demokritos" Athens15341 Greece

Josephson junctions (JJs) are by nature neuromorphic hardware devices capable of mimicking excitability and spiking dynamics. When coupled together or combined with other superconducting elements, they can emulate additional behaviors found in biological neurons. From a technological point of view, JJ-based neuromorphic systems are particularly appealing since they present THz-speed processing and they operate with near-zero power dissipation. In this work we study a system of inductively coupled JJs and focus on the nonlinear dynamical aspects of its neurocomputational properties. In particular, we report on spiking behavior related to a saddle-node off invariant cycle bifurcation and excitability type II, synchronization, first spike latency effects, and multistability. Special emphasis is placed on the bursting dynamics the system is capable of reproducing, and a new underlying mechanism is proposed beyond the approach followed in prior works. Copyright © 2025, The Authors. All rights reserved.

关键词： Tunnel junctions

来源：评论

学校读者我要写书评

暂无评论

A Priority-Aware AI-Generated Content Resource Allocation Method for Multi-UAV Aided Metaverse

A Priority-Aware AI-Generated Content Resource Allocation Me...

引用

IEEE Conference on Wireless Communications and Networking

作者： Zheng Zhang Jingjing Wang Jianrui Chen Zhengru Fang Chunxiao Jiang Zhu Han School of Cyber Science and Technology Beihang University Beijing China Department of Computer Science City University of Hong Kong Hong Kong China Beijing National Research Center for Information Science and Technology Tsinghua University Beijing China Department of Electrical and Computer Engineering University of Houston Houston TX USA

ISBN: (数字)9798350368369

ISBN: (纸本)9798350368376

With the advancement of large model technologies, AI -generated content is gradually emerging as a mainstream method for content creation. The metaverse, as a key application scenario for the next-generation communication technologies, heavily depends on advanced content generation technologies. Nevertheless, the diverse types of metaverse applications and their stringent real-time requirements constrain the full potential of AIGC technologies within this environment. In order to tackle with this problem, we construct a priority-aware multi-UAV aided metaverse system and formulate it as a Markov decision process (MDP). We propose a diffusion-based reinforcement learning algorithm to solve the resource allocation problem and demonstrate its superiority through enough comparison and ablation experiments.

关键词： Metaverse Simulation Markov decision processes Reinforcement learning Autonomous aerial vehicles Diffusion models Real-time systems Communications technology Resource management Next generation networking

来源：评论

学校读者我要写书评

暂无评论

Service Migration for Delay-Sensitive IoT Applications in Edge Networks

引用

IEEE Transactions on Services Computing 2025年

作者： Li, Xiaocui Zhou, Zhangbing Wang, Yasha Deng, Shuiguang Hung, Patrick C. K. School of Information Engineering Beijing100083 China Peking University National Research & Engineering Center of Software Engineering Beijing100871 China Peking University New Generation Information Technology Research Institute Tianjin300380 China School of Information Engineering Frontiers Science Center for Deep-Time Digital Earth Beijing100083 China Institut Polytechnique de Paris Télécom SudParis Palaiseau91120 France Zhejiang University College of Computer Science and Technology Hangzhou310012 China Ontario Tech University Faculty of Business and Information Technology Oshawa Canada

The proliferation of Internet of Things (IoT) applications prompts extraordinary demands for the collaboration of large amounts of computational resources provided by IoT devices in edge networks, and these applications are mostly delay-sensitive. Generally, these resources are encapsulated as IoT services. Thereafter, IoT applications can be performed, such that the collaboration of their sub-tasks is achieved through the composition of functionally complementary and geographically contiguous IoT services. The status of computational resources in IoT devices may change continuously along with their occupancy and release by IoT services. Considering the resource-scarceness of IoT devices, when the workload of IoT devices increases due to more services to be processed, certain IoT devices may hardly have enough remaining resources to co-host more instances of certain IoT services prescribed by forthcoming IoT applications with strict constraints. As a result, the delay satisfaction of both on-running and forthcoming IoT applications may be negatively impacted, or even hardly be satisfied any longer. To solve this issue, this paper proposes a rEsource-Efficient service Configuration (E2rC) mechanism, which aims to optimize the configuration of computational resources provided by IoT devices with respect to complex requirements prescribed by IoT applications, through service migration techniques. This service migration problem is formulated as markov multi-phases decisions, which is solved through our enhanced Deep Reinforcement Learning (DRL) approach with a two-layer Q-network. Extensive experiments have been conducted upon the dataset of our testbed system. Evaluation results show that our E2rC is more efficient than the state-of-art counterparts in satisfying delay constraints of IoT applications, while reducing the energy consumption and improving the resource utilization efficiency of IoT devices. © 2025 IEEE. All rights reserved.

关键词： Markov processes

来源：评论

学校读者我要写书评

暂无评论

Correction: A new quantum-enhanced approach to AI-driven medical imaging system

引用

Cluster Computing 2025年第4期28卷 1-3页

作者： Ahmadpour, Seyed-Sajad Avval, Danial Bakhshayeshi Darbandi, Mehdi Navimipour, Nima Jafari Ain, Noor Ul Kassa, Sankit Department of Computer Engineering Faculty of Engineering and Natural Sciences Kadir Has University Istanbul Türkiye Department of Information Systems Engineering Sakarya University Sakarya Türkiye Department of Electrical and Electronic Engineering Eastern Mediterranean University Gazimagusa Türkiye Future Technology Research Center National Yunlin University of Science and Technology Douliou Taiwan Research Center of High Technologies and Innovative Engineering Western Caspian University Baku Azerbaijan Department of Business Administration Kadir Has University Istanbul Türkiye Department of Electronics and Telecommunication Symbiosis Institute of Technology Symbiosis International (Deemed University) Pune India

来源：评论

学校读者我要写书评

暂无评论

Energy Oriented Three-Tier Computation Offloading Scheme in Maritime Edge Computing Network

引用

IEEE Transactions on Vehicular technology 2025年

作者： Zhang, Hongxia Xi, Shiyu Shang, Bodong Zhang, Peiying Wu, Sheng Jiang, Chunxiao Qingdao Institute of Software College of Computer Science and Technology Qingdao266580 China Eastern Institute of Technology Eastern Institute for Advanced Study Ningbo315200 China Beijing University of Posts and Telecommunications School of Information and Communication Engineering Beijing100876 China Tsinghua University Beijing National Research Center for Information Science and Technology Beijing100084 China

With the development of maritime continuing to grow, large-scale maritime wireless devices (MWDs) are being deployed for various maritime applications. The rapid development of maritime Internet of Things (IoT) and growing demand for these applications have led to an increasing number of computation-intensive and resource-sensitive tasks. Considering the limited computing capacity and energy of MWDs, it is challenging to transmit and process massive maritime computing tasks in an energy efficient manner. In this paper, we propose a three-tier maritime edge computing system in which low earth orbit (LEO) satellite and offshore base station (OBS) support to provide communication and computing services for MWDs. The system optimizes association variable, task partitioning, transmission power, and computing resource allocation to reduce system energy consumption under latency constraints. The optimization problem is formulated as a mixed integer non-linear programming (MINLP) optimization problem. We decompose the original problem into four subproblems and design algorithms to solve them. Firstly, the association variable optimization subproblem is solved by slack variable method. Then, we adopt quadratic transformation and design a difference of convex algorithm to solve transmission power optimization subproblem. Further, task partitioning subproblem is solved by obtaining upper and lower bounds of offloading task size by strict deduction, and then solved by the standard convex method. Finally, Lagrangian dual method is utilized to deal with joint computing resource optimization subproblem of LEO satellite and OBS. In simulation results, the proposed algorithm saves 39.3% of system energy consumption compared to benchmark schemes. © 1967-2012 IEEE.

关键词： Computation offloading

来源：评论

学校读者我要写书评

暂无评论

Geometric Feature-Driven Metric Learning for 3D Craniofacial Superimposition

Geometric Feature-Driven Metric Learning for 3D Craniofacial...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Qingdong Long Junli Zhao Fuqing Duan Chengyuan Wang Xuesong Wang Lijie Geng Zhenkuan Pan Mingquan Zhou College of Computer Science and Technology Qingdao University Qingdao China School of Artificial Intelligence Beijing Normal University Beijing China National and Local Joint Engineering Research Center for Cultural Heritage Digitization School of Information Science and Technology Northwest University Xi’an China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Craniofacial superimposition is a crucial forensic science technique to identify human remains by matching skulls to facial images. However, this task is challenging due to significant morphological differences between skulls and faces, limited paired samples, and high data dimensionality. We proposed a geometric feature-driven metric learning method for craniofacial superimposition to address these issues. Firstly, we extracted geometric features, including depth, curvature, and elevation of 3D craniofacial data, to generate 2D maps of structured representations enriched with geometric details. Next, we novelly designed a Triplet Network for geometric feature-driven metric learning, which leverages triplet loss to learn discriminative embeddings and effectively handle the limited paired data problem. By incorporating the Sinkhorn Distance as an additional constraint, we aligned the skull and face data distributions, enhancing the matching precision. We conducted extensive experiments on a 3D craniofacial dataset, achieving a maximum accuracy of 99.45% on curvature maps, surpassing state-of-the-art methods. Our code will be available after publication at https://***/Lqd-js/cranial-superimposition.

关键词： Measurement Learning systems Three-dimensional displays Accuracy Forensics Signal processing Feature extraction Robustness Speech processing Faces

来源：评论

学校读者我要写书评

暂无评论

Enhancing Diagnostic Accuracy for Knee Osteoarthritis Using a Deep Learning Approach

Enhancing Diagnostic Accuracy for Knee Osteoarthritis Using ...

引用

International Symposium on Applied Science 2024, ISAS 2024

作者： Raza, Aquib Ching, Congo Tak Shing Liu, Peng-Ta Phan, Thien-Luan Tran, Ngoc-Luan Nguyen, Ha Anh T. Rameez, Mohammad Li, Hung-Chung Linh, Dao Nguyen Phuong Van Hieu, Nguyen Tran, Trung Nghia National Taiwan University Taipei City10673 Taiwan Research Center for Information Technology Academia Sinica Taipei City11529 Taiwan Graduate Institute of Biomedical Engineering National Chung Hsing University Taichung402 Taiwan Department of Electrical Engineering National Chi Nan University Nantou County Taiwan Advanced Plant and Food Crop Biotechnology Center National Chung Hsing University Taichung402 Taiwan International Doctoral Program in Agriculture National Chung Hsing University Taichung402 Taiwan Fall Prevention Center Department of Physical Medicine & Rehabilitation Changhua Christian Hospital Changhua500 Taiwan Department of Physics and Electronic Engineering University of Science Vietnam National University - Ho Chi Minh City Ho Chi Minh City70000 Viet Nam Department of Computer Science University of York York HeslingtonYO10 5DD United Kingdom Undergraduate Program of Intellectual Creativity Engineering National Chung Hsing University Taichung402 Taiwan Ho Chi Minh City72409 Viet Nam HCMUT Vietnam National University Ho Chi Minh City Thu Duc Ho Chi Minh City71308 Viet Nam

This research introduces a deep learning framework that combines convolutional neural networks with autoencoders to improve the diagnostic accuracy of knee osteoarthritis. The study utilized a publicly available dataset from the Mendeley Data repository, where knee joint images were preprocessed using Contrast Limited Adaptive Histogram Equalization to enhance contrast and ensure consistent image sizing. Initially, an autoencoder was employed in an unsupervised manner to learn compressed representations of the data. These representations were then used to train a CNN to classify knee images into 'healthy' and 'osteoarthritic' categories. The model was trained using an 80:20 split for training and testing data, incorporating early stopping and data augmentation techniques to prevent overfitting and enhance generalization. Performance metrics such as accuracy, precision, recall, and the area under the Receiver Operating Characteristic curve (ROCAUC) were used to evaluate the model, resulting in a high test accuracy of 98.95%. The high precision and recall rates further confirmed the model's robustness in diagnosing KOA. To enhance interpretability, Gradient-weighted Class Activation Mapping (Grad-CAM) was applied to identify critical regions in the images that influenced the model's decisions. Additionally, t-Distributed Stochastic Neighbor Embedding (t-SNE) was used to visualize the high-dimensional encoded features in a two-dimensional space, providing deeper insights into how the model differentiates between classes. The promising results of this deep learning approach suggest its potential for broader applications in medical imaging diagnostics. By providing healthcare professionals with a tool that enhances diagnostic accuracy and operational efficiency, this framework lays the foundation for developing scalable, AI-driven diagnostic systems for various complex medical conditions. © 2025 Institute of Physics Publishing. All rights reserved.

关键词： Joints (anatomy)

来源：评论

学校读者我要写书评

暂无评论

A Joint Visual Compression and Perception Framework for Neuralmorphic Spiking Camera

arXiv

引用

arXiv 2025年

作者： Feng, Kexiang Jia, Chuanmin Ma, Siwei Gao, Wen Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China National Engineering Research Center of Visual Technology Peking University Beijing100871 China Wangxuan Institute of Computer Technology Peking University Beijing100080 China

The advent of neuralmorphic spike cameras has garnered significant attention for their ability to capture continuous motion with unparalleled temporal resolution. However, this imaging attribute necessitates considerable resources for binary spike data storage and transmission. In light of compression and spike-driven intelligent applications, we present the notion of Spike Coding for Intelligence (SCI), wherein spike sequences are compressed and optimized for both bit-rate and task performance. Drawing inspiration from the mammalian vision system, we propose a dual-pathway architecture for separate processing of spatial semantics and motion information, which is then merged to produce features for compression. A refinement scheme is also introduced to ensure consistency between decoded features and motion vectors. We further propose a temporal regression approach that integrates various motion dynamics, capitalizing on the advancements in warping and deformation simultaneously. Comprehensive experiments demonstrate our scheme achieves state-of-the-art (SOTA) performance for spike compression and analysis. We achieve an average 17.25% BD-rate reduction compared to SOTA codecs and a 4.3% accuracy improvement over SpiReco for spike-based classification, with 88.26% complexity reduction and 42.41% inference time saving on the encoding side. Copyright © 2025, The Authors. All rights reserved.

关键词： Image compression

来源：评论

学校读者我要写书评

暂无评论

Source-free Semantic Regularization Learning for Semi-supervised Domain Adaptation

arXiv

引用

arXiv 2025年

作者： Huang, Xinyang Zhu, Chuang Ren, Ruiying Liu, Shengjie Huang, Tiejun The School of Artificial Intelligence Beijing University of Posts and Telecommunications Beijing China The BAAI Institute for Artificial Intelligence Peking University Beijing China The National Engineering Research Center of Visual Technology School of Computer Science Peking University China

Semi-supervised domain adaptation (SSDA) has been extensively researched due to its ability to improve classification performance and generalization ability of models by using a small amount of labeled data on the target domain. However, existing methods cannot effectively adapt to the target domain due to difficulty in fully learning rich and complex target semantic information and relationships. In this paper, we propose a novel SSDA learning framework called semantic regularization learning (SERL), which captures the target semantic information from multiple perspectives of regularization learning to achieve adaptive fine-tuning of the source pre-trained model on the target domain. SERL includes three robust semantic regularization techniques. Firstly, semantic probability contrastive regularization (SPCR) helps the model learn more discriminative feature representations from a probabilistic perspective, using semantic information on the target domain to understand the similarities and differences between samples. Additionally, adaptive weights in SPCR can help the model learn the semantic distribution correctly through the probabilities of different samples. To further comprehensively understand the target semantic distribution, we introduce hard-sample mixup regularization (HMR), which uses easy samples as guidance to mine the latent target knowledge contained in hard samples, thereby learning more complete and complex target semantic knowledge. Finally, target prediction regularization (TPR) regularizes the target predictions of the model by maximizing the correlation between the current prediction and the past learned objective, thereby mitigating the misleading of semantic information caused by erroneous pseudo-labels. Extensive experiments on three benchmark datasets demonstrate that our SERL method achieves state-of-the-art performance. © 2025, CC BY.

关键词： Latent semantic analysis

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：