检索结果-内蒙古大学图书馆

Domain adaptive FL for edge-enabled privacy-preserving MRI analysis

Alexandria engineering Journal 2025年 128卷 324-339页

作者： Saeed Iqbal Imran Arshad Choudhry Inam Ullah Kuljeet Kaur Bong Jun Choi Mohammad Mehedi Hassan College of Mechatronics and Control Engineering Shenzhen University Shenzhen 518060 China Department of Computer Science Faculty of Information Technology & Computer Science University of Central Punjab Lahore Pakistan Department of Computer Engineering Gachon University Seongnam 13120 Republic of Korea Electrical Engineering Department École de technologie supérieure Montréal QC H3C1K3 Canada Centre for Research Impact & Outcome Chitkara University Institute of Engineering and Technology Chitkara University Rajpura 140401 Punjab India School of Computer Science and Engineering Soongsil University Seoul Republic of Korea Department of Information Systems College of Computer and Information Sciences King Saud University Riyadh 11543 Saudi Arabia

Data heterogeneity, privacy leakage challenges, the ineffectiveness of conventional collaborative learning techniques, and unresolved managing non-IID data distributions are some of the major obstacles to implementing artificial intelligence (AI) in healthcare. Federated learning (FL) frameworks frequently have trouble distinguishing between privacy protection and model accuracy, especially when used for delicate medical imaging applications. This study presents a novel framework that synergizes federated learning (FL) with edge computing to address these issues while safeguarding patient privacy. Our proposed Domain Adaptive Federated (DAD) learning approach effectively mitigates both inter-client and intra-client data heterogeneity, enabling collaborative model training across diverse medical imaging modalities (MRI, CT, PET) through cross-domain adaptation. Experimental evaluations on MRI brain segmentation datasets demonstrate the superior performance of DAD compared to traditional FL methods, as evidenced by significant improvements in F1-score (96.3), sensitivity (96.0), specificity (97.1), and AUC (96.7). This enhanced accuracy and robustness in handling heterogeneous and privacy-sensitive data render DAD an ideal candidate for privacy-preserving AI in consumer healthcare. By pioneering innovative strategies for collaborative model training and data privacy, this research contributes to the emerging field of edge intelligence, paving the way for improved patient outcomes while adhering to stringent confidentiality and ethical mandates.

关键词： Multi-modal data Domain adaptive learning Federated learning Privacy preservation Domain-specific Data privacy

来源：评论

学校读者我要写书评

暂无评论

Data Heterogeneity Challenge in Consumer Electronics: A Federated Learning Solution

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Liu, Zhenyu Wu, Yunting Ge, Huanhuan Wang, Jinlong Lyu, Zhihan Qingdao University of Technology School of Information and Control Engineering Shandong Qingdao China School of Information and Control Engineering Innovation Institute for Sustainable Maritime Architecture Research and Technology Shandong Qingdao China School of Computer Science and Technology Xidian University Xi’an710126 China

Federated learning(FL) enables collaborative model training across consumer electronic devices while preserving data privacy. However, the non-independent and identically distributed nature of data in real-world scenarios significantly impacts personalization performance. To address this challenge, we propose pFedPKD, in this framework, we first employ model pruning to simplify the global model, thereby enhancing its adaptability and improving the efficiency of knowledge transfer between different electronics devices. Moreover, we introduce an attention-based knowledge transfer mechanism, which leverages knowledge distillation to transfer the attention map and predictions of the pruned global model to the local models, thus improving their generalization capability and performance. Extensive experiments on three datasets demonstrate that pFedPKD consistently outperforms state-of-the-art methods, achieving higher personalization accuracy and robustness under heterogeneous data distributions. This work provides an efficient solution for applying FL in consumer electronics, offering improved adaptability and resource efficiency. © 1975-2011 IEEE.

关键词： Knowledge transfer

来源：评论

学校读者我要写书评

暂无评论

MedFILIP: Medical Fine-grained Language-Image Pre-training

arXiv

引用

arXiv 2025年

作者： Liang, Xinjie Li, Xiangyu Li, Fanding Jiang, Jie Dong, Qing Wang, Wei Wang, Kuanquan Dong, Suyu Luo, Gongning Li, Shuo School of Computer Science and Technology Harbin Institute of Technology Harbin China Department of Thoracic Surgery at No. 4 Affiliated Hospital Harbin Medical University Harbin China School of Computer Science and Technology Harbin Institute of Technology Shenzhen China College of computer and control engineering Northeast Forestry University Harbin China Computer Electrical and Mathematical Sciences & Engineering Division King Abdullah University of Science and Technology Thuwal Saudi Arabia Department of Biomedical Engineering Department of Computer and Data Science Case Western Reserve University ClevelandOH United States

Medical vision-language pretraining (VLP) that leverages naturally-paired medical image-report data is crucial for medical image analysis. However, existing methods struggle to accurately characterize associations between images and diseases, leading to inaccurate or incomplete diagnostic results. In this work, we propose MedFILIP, a fine-grained VLP model, introduces medical image-specific knowledge through contrastive learning, specifically: 1) An information extractor based on a large language model is proposed to decouple comprehensive disease details from reports, which excels in extracting disease deals through flexible prompt engineering, thereby effectively reducing text complexity while retaining rich information at a tiny cost. 2) A knowledge injector is proposed to construct relationships between categories and visual attributes, which help the model to make judgments based on image features, and fosters knowledge extrapolation to unfamiliar disease categories. 3) A semantic similarity matrix based on fine-grained annotations is proposed, providing smoother, information-richer labels, thus allowing fine-grained image-text alignment. 4) We validate MedFILIP on numerous datasets, e.g., RSNA-Pneumonia, NIH ChestX-ray14, VinBigData, and COVID-19. For single-label, multi-label, and fine-grained classification, our model achieves state-of-the-art performance, the classification accuracy has increased by a maximum of 6.69%. The code is available in https://***/ PerceptionComputingLab/MedFILIP. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Innate Cyber-Immunity and Acquired Cyber-Immunity across Multi-Area Power Systems

引用

IEEE Transactions on Power Systems 2025年

作者： Hou, Jiazuo Song, Yue Hou, Yunhe Peng, Jimmy Chih-Hsien National University of Singapore Department of Electrical and Computer Engineering Singapore Tongji University Department of Control Science and Engineering National Key Laboratory of Autonomous Intelligent Unmanned Systems Shanghai Institute of Intelligent Science and Technology Shanghai201804 China The University of Hong Kong Department of Electrical and Electronic Engineering Hong Kong University of Hong Kong Shenzhen Institute of Research and Innovation Shenzhen518057 China

Organisms have evolved innate and acquired immune systems to defend against pathogens like coronaviruses. Similarly, power networks, threatened by cyber-attacks, desire cyber-immunity. Inspired by the immunology research, this study develops a cyber-immune strategy-characterizing innate cyber-immunity and establishing acquired cyber-immunity-for any selected portion of a power network (termed as power subnetwork) against stealthy false data injection (FDI) cyberattacks. Therein, without extrinsic protection and information asymmetry, this paper for the first time confirms the existence and provides closed-form conditions for power subnetworks that are innately immune to FDI cyber-attacks despite all measurements being compromised. Single-leader-multi-follower bi-level optimization models are then developed to strategically allocate measurement protection, leading to the acquired cyber-immunity, whose conditions for existence and the least cost are explicitly formulated. More importantly, a bi-directional cyber-immunity propagation mechanism is presented to provide the optimal spatial allocation of defense resources across adjacent subnetworks. It also explicitly determines the temporal allocation sequence, revealing its sequential and cumulative impacts on establishing cyber-immunity. This is achieved by designing recursive traversal algorithms, one of which is modelled as a post-order depth-first-search problem of a multi-fork tree. The proposed subnetwork-specific cyber-immune strategy is verified in a modified Australian National Electricity Market. © 1969-2012 IEEE.

关键词： Phishing

来源：评论

学校读者我要写书评

暂无评论

A Survey of fMRI to Image Reconstruction

arXiv

引用

arXiv 2025年

作者： Guo, Weiyu Sun, Guoying He, Jianxiang Shao, Tong Wang, Shaoguang Chen, Ziyang Hong, Meisheng Sun, Ying Xiong, Hui Thrust of Artificial Intelligence The Hong Kong University of Science and Technology Guangzhou China College of Computer Science and Technology Harbin Institute of Technology Shenzhen China School of Control Science and Engineering Shandong University China

Functional magnetic resonance imaging (fMRI)based image reconstruction plays a pivotal role in decoding human perception, with applications in neuroscience and brain-computer interfaces. While recent advancements in deep learning and large-scale datasets have driven progress, challenges such as data scarcity, cross-subject variability, and low semantic consistency persist. To address these issues, we introduce the concept of fMRI-to-Image Learning (fMRI2Image) and present the first systematic review in this field. This review highlights key challenges, categorizes methodologies such as fMRI signal encoding, feature mapping, and image generator. Finally, promising research directions are proposed to advance this emerging frontier, providing a reference for future studies. © 2025, CC BY.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Optimal Placement of Nature-Based Solutions for Urban Challenges

arXiv

引用

arXiv 2025年

作者： Pinto, Diego Maria Russo, Davide Donato Sudoso, Antonio M. Institute for Systems Analysis and Computer Science "Antonio Ruberti" National Research Council Via dei Taurini 19 Rome00185 Italy Department of Computer Control and Management Engineering "Antonio Ruberti" Sapienza University of Rome Via Ariosto 25 Rome00185 Italy

Increased urbanization and climate change intensify urban heat islands and degrade air quality, making current mitigation strategies insufficient. Nature-based solutions (NBSs), such as parks, green walls, roofs, and street trees, offer a promising means to regulate urban temperatures and enhance air quality. However, determining their optimal placement to maximize environmental benefits remains a pressing challenge. Leveraging Operational Research (OR) tools, we propose a Mixed-Integer Linear Programming (MILP) model that integrates multiple factors, including urban challenges, physical constraints, clustering techniques, convolution theory, and fairness considerations. This model determines the optimal placement of NBSs by addressing metrics such as ground temperature, air quality, and accessibility to green spaces. Through several case study analyses, we demonstrate the effectiveness of our approach in improving environmental and social indicators. This research holds implications for policy and practice, empowering urban planners and policymakers to make informed decisions regarding NBS implementation. Such decisions ensure that investments in urban greening yield maximum environmental, social, and economic benefits. Copyright © 2025, The Authors. All rights reserved.

关键词： Mixed-integer linear programming

来源：评论

学校读者我要写书评

暂无评论

Human-centric assembly in smart factories

引用

CIRP Annals 2025年

作者： Wang, Lihui Gao, Robert X. Krüger, Jörg Váncza, József Department of Production Engineering KTH Royal Institute of Technology Stockholm Sweden Department of Mechanical and Aerospace Engineering Case Western Reserve University Cleveland United States Institute for Machine Tools and Factory Management Technical University Berlin Germany Laboratory of Engineering and Management Intelligence Institute for Computer Science and Control Hungary

Assembly in future smart factories needs to address three challenges, including human centricity, sustainability, and resilience. Conventional approaches for automation in assembly have reached a bottleneck in terms of operation automomy, leaving various tasks to continued manual labour by human operators. To ease the burden on humans both physically and intellectually, human-centric assembly enhanced by augmented robots, cognitive systems, mixed reality and collaborative intelligence, assisted by thought-driven brain robotic controls, provides a promising solution. Within the context, this keynote provides an in-depth analysis of the state of human-centric assembly and identifies potentially fruitful research directions in future smart factories. © 2025 The Author(s)

关键词： Factory automation

来源：评论

学校读者我要写书评

暂无评论

Finite Sample Analysis of Subspace Identification Methods

arXiv

引用

arXiv 2025年

作者： He, Jiabao Ziemann, Ingvar Rojas, Cristian R. Qin, S. Joe Hjalmarsson, Håkan Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm100 44 Sweden University of Pennsylvania PhiladelphiaPA19104 United States Institute of Data Science Lingnan University Hong Kong

As one of the mainstream approaches in system identification, subspace identification methods (SIMs) are known for their simple parameterization for MIMO systems and robust numerical properties. However, a comprehensive statistical analysis of SIMs remains an open problem. Amid renewed focus on identifying state-space models in the non-asymptotic regime, this work presents a finite sample analysis for a large class of open-loop SIMs. It establishes high-probability upper bounds for system matrices obtained via SIMs, and reveals that convergence rates for estimating Markov parameters and system matrices are O(1/√N) up to logarithmic terms, in line with classical asymptotic results. Following the key steps of SIMs, we arrive at the above results by a three-step procedure. In Step 1, we begin with a parsimonious SIM (PARSIM) that uses least-squares regression to estimate multiple high-order ARX models in parallel. Leveraging a recent analysis of an individual ARX model, we obtain a union error bound for a bank of ARX models. Step 2 involves model reduction via weighted singular value decomposition (SVD), where we consider different data-dependent weighting matrices and use robustness results for SVD to obtain error bounds on extended controllability and observability matrices, respectively. The final Step 3 focuses on deriving error bounds for system matrices, where two different realization algorithms, the MOESP type and the Larimore type, are considered. Although our study initially focuses on PARSIM, the methodologies apply broadly across many variants of SIMs. Copyright © 2025, The Authors. All rights reserved.

关键词： MIMO systems

来源：评论

学校读者我要写书评

暂无评论

Multi-Agent Distributed DQN and Transfer Learning for Energy-Efficient Power Management in Solar Energy-Harvested Small-Cell Networks

引用

IEEE Internet of Things Journal 2025年

作者： Cho, Hyebin Kim, Hyungsub Na, JeeHyeon Lim, Seung-Chan Lee, Howon Hankyong National University School of Electrical Electronic and Control Engineering Anseong17579 Korea Republic of Electronics and Telecommunications Research Institute Daejeon34141 Korea Republic of Hankyong National University School of ICT Robotics and Mechanical Engineering and IITC Anseong17579 Korea Republic of Ajou University Department of Electrical and Computer Engineering Suwon16499 Korea Republic of

The integration of solar energy harvesting into small-cell networks is a promising solution for achieving energy-efficient and sustainable wireless communications. However, the inherent variability and intermittency of solar energy, coupled with precise inter-cell interference management, significantly hinder efficient network operation. To resolve these challenges, we propose a multi-agent distributed deep Q-network (MA-DDQN) framework, where the distributed base stations learn the optimal transmit power control policies. To further enhance adaptability under varying solar conditions, we present a daily model transfer with a fine-tuning approach, enabling efficient deployment without extensive training overhead. Simulation results demonstrate that the proposed methods remarkably improve energy efficiency while maintaining robust adaptability under dynamic solar conditions, revealing their potential for sustainable small-cell network deployments. © 2014 IEEE.

关键词： Solar cell efficiency

来源：评论

学校读者我要写书评

暂无评论

CorPN: Corner Prediction Network for Visual Tracking

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Li, Qun Zhang, Haijun Yang, Kai Dai, Wei Ma, Jianghong Liu, Yuping the Department of Computer Science Harbin Institute of Technology Shenzhen Xili University Town Shenzhen518055 China the School of Computer Science and Artificial Intelligence WuHan Textile University Wuhan430200 China the School of Information and Control Engineering Artificial Intelligence Research Institute China University of Mining and Technology Xuzhou221116 China the School of Microelectronics Fudan University Shanghai200433 China

Object tracking is a well-established task in the field of computer vision. Despite numerous efforts over the last few decades, object tracking remains a formidable challenge, largely due to the complexity of the scenes encountered in video sequences. In line with the powerful capabilities of Transformers for object feature extraction and fusion, as well as drawing inspiration from CornerNet, this paper introduces a new Transformer-based anchor-free method designed for object tracking called CorPN (i.e., Corner Prediction Network). Specifically, a Transformer-based backbone is proposed to extract and fuse deep features in a Siamese network architecture. Furthermore, our methodology aims to emulate the intuitive cognitive process by which humans identify the bounding-box corners of objects. CorPN turns out to be a more effective and robust tracking method because it jointly leverages multi-level feature aggregation, corner prediction, and probabilistic prediction incorporated in the tracking head. In extensive evaluations on public datasets, our method achieves a state-of-the-art performance. In particular, our approach achieves an area-under-the-curve (AUC) of 85.2%, a precision score (P) of 84.5%, and a normalized precision score (Pnorm) of 89.6%, outperforming extant state-of-the-art trackers on these metrics. © 2025 IEEE. All rights reserved,

关键词： Object 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：