检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Han, Pengchao Shi, Xingyan Huang, Jianwei The School of Information Engineering Guangdong University of Technology Guangzhou510006 China The School of Data Science The Chinese University of Hong Kong Shenzhen China The School of Science and Engineering Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen Key Laboratory of Crowd Intelligence Empowered Low-Carbon Energy Network CSIJRI Joint Research Centre on Smart Energy Storage The Chinese University of Hong Kong Guangdong Shenzhen518172 China

Knowledge distillation (KD) can enable collaborative learning among distributed clients that have different model architectures and do not share their local data and model parameters with others. Each client updates its local model using the average model output/feature of all client models as the target, known as federated KD. However, existing federated KD methods often do not perform well when clients’ local models are trained with heterogeneous local datasets. In this paper, we propose Federated knowledge distillation enabled by Adversarial Learning (FedAL) to address the data heterogeneity among clients. First, to alleviate the local model output divergence across clients caused by data heterogeneity, the server acts as a discriminator to guide clients’ local model training to achieve consensus model outputs among clients through a min-max game between clients and the discriminator. Moreover, catastrophic forgetting may happen during the clients’ local training and global knowledge transfer due to clients’ heterogeneous local data. Towards this challenge, we design the less-forgetting regularization for both local training and global knowledge transfer to guarantee clients’ ability to transfer/learn knowledge to/from others. Experimental results show that FedAL and its variants achieve higher accuracy than other federated KD baselines. Copyright © 2023, The Authors. All rights reserved.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Exact Fit Attention in Node-Holistic Graph Convolutional Network for Improved EEG-Based Driver Fatigue Detection

arXiv

引用

arXiv 2025年

作者： Xu, Meiyan Chen, Qingqing Chen, Duo Ding, Yi Wang, Jingyuan Gu, Peipei Pan, Yijie Huang, Deshuang Guo, Jiayang Zhang, Xun Department of Computer Science The Institute of Human Factors Engineering Minnan Normal University Zhangzhou363000 China School of Artificial Intelligence and Information Technology Nanjing University of Chinese Medicine Nanjing210023 China College of Computing and Data Science Nanyang Technological University Singapore School of Digital Media and Design Arts Beijing University of Posts and Telecommunications Beijing100088 China College of Software Engineering Zhengzhou University of Light Industry Zhengzhou450000 China Department of Computer Science and Technology Tsinghua University Beijing100084 China Eastern Institute for Advanced Study Eastern Institute of Technology Ningbo315200 China Ningbo Institute of Digital Twin Eastern Institute of Technology Ningbo315200 China National Institute for Data Science in Health and Medicine Xiamen University Xiamen361005 China

EEG-based fatigue monitoring can effectively reduce the incidence of related traffic accidents. In the past decade, with the advancement of deep learning, convolutional neural networks (CNN) have been increasingly used for EEG signal processing. However, due to the data’s non-Euclidean characteristics, existing CNNs may lose important spatial information from EEG, specifically channel correlation. Thus, we propose the node-holistic graph convolutional network (NHGNet), a model that uses graphic convolution to dynamically learn each channel’s features. With exact fit attention optimization, the network captures inter-channel correlations through a trainable adjacency matrix. The interpretability is enhanced by revealing critical areas of brain activity and their interrelations in various mental states. In validations on two public datasets, NHGNet outperforms the SOTAs. Specifically, in the intra-subject, NHGNet improved detection accuracy by at least 2.34% and 3.42%, and in the inter-subjects, it improved by at least 2.09% and 15.06%. Visualization research on the model revealed that the central parietal area plays an important role in detecting fatigue levels, whereas the frontal and temporal lobes are essential for maintaining vigilance. Copyright © 2025, The Authors. All rights reserved.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

SVM-based Predictive Modeling of Drowsiness in Hospital Staff for Occupational Safety Solution via IoT Infrastructure

SVM-based Predictive Modeling of Drowsiness in Hospital Staf...

引用

Computer, Communication and Control (IC4), International Conference on

作者： B. Jai Ganesh P. Vijayan V. Vaidehi S. Murugan R. Meenakshi M. Rajmohan Department of Electronics and Communication Engineering R.M.D. Engineering College Chennai Tamil Nadu India Department of Artificial Intelligence and Data Science Saveetha Engineering College Chennai Tamil Nadu India Department of Computer Science Vinayaka Mission’s Kirupananda Variyar Arts and Science College Vinayaka Mission’s Research Foundation (DU) Salem Tamil Nadu India Department of Biomedical Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Saveetha University Chennai Tamil Nadu India Department of Computer Science and Engineering Chennai Institute of Technology Chennai Tamil Nadu India Department of Electronics and Communication Engineering Hindustan Institute of Technology and Science Chennai Tamil Nadu India

This research proposes an efficient approach to improve hospital worker safety on the job by detecting and preventing drowsiness. The system incorporates data from hospital-deployed Internet of Things (IoT) infrastructure and using Support Vector Machines (SVM) as its primary predictive modeling tool. data pertaining to ambient factors, work habits, and physiological indications of employees are collected in real-time by the IoT infrastructure. To identify patterns linked to sleepiness, the SVM model is trained on an extensive dataset. The next step is to implement the model in a healthcare setting so that it can detect signs of sleepiness and act accordingly in a timely manner. With the help of the IoT the predictive model can be dynamically adjusted and monitored in real-time, making it useful in a wide range of situations. The recommended approach is to enhance workplace safety by reducing the likelihood of accidents caused by sleepiness among healthcare workers. The SVM-based prediction model can correctly detect and forecast cases of sleepiness, the research subjected it to extensive testing and validation. This system's adoption helps with the creation of data-driven, proactive methods to improve hospital occupational safety.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Low-Latency Layer-Aware Proactive and Passive Container Migration in Meta Computing

arXiv

引用

arXiv 2024年

作者： Liu, Mengjie Li, Yihua Mou, Fangyi Tang, Zhiqing Lou, Jiong Guo, Jianxiong Jia, Weijia Institute of Artificial Intelligence and Future Networks Beijing Normal University China Faculty of Arts and Sciences Beijing Normal University China Guangdong Key Lab of AI and Multi-Modal Data Processing BNU-HKBU United International College China Department of Computer Science and Engineering Shanghai Jiao Tong University China Department of Informatics University of Zurich Zurich Switzerland

关键词： Containers

来源：评论

学校读者我要写书评

暂无评论

NTIRE 2023 Challenge on Image Super-Resolution (×4): Methods and Results

NTIRE 2023 Challenge on Image Super-Resolution (×4): Method...

引用

2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023

作者： Zhang, Yulun Zhang, Kai Chen, Zheng Li, Yawei Timofte, Radu Zhang, Junpei Zhang, Kexin Peng, Rui Ma, Yanbiao Jiao, Licheng Huang, Huaibo Zhou, Xiaoqiang Ai, Yuang He, Ran Qiu, Yajun Zhu, Qiang Li, Pengfei Li, Qianhui Zhu, Shuyuan Zhang, Dafeng Li, Jia Wang, Fan Li, Chunmiao Kim, TaeHyung Kil, Jungkeong Kim, Eon Yu, Yeonseung Lee, Beomyeol Lee, Subin Lim, Seokjae Chae, Somi Choi, Heungjun Huang, ZhiKai Chen, YiChung Chiang, YuanChun Yang, HaoHsiang Chen, WeiTing Chang, HuaEn Chen, I-Hsiang Hsieh, ChiaHsuan Kuo, SyYen Choi, Ui-Jin Conde, Marcos V. Khowaja, Sunder Ali Yoon, Jiseok Lee, Ik Hyun Gendy, Garas Sabor, Nabil Hou, Jingchao He, Guanghui Zhang, Zhao Li, Baiang Zheng, Huan Zhao, Suiyi Gao, Yangcheng Wei, Yanyan Ren, Jiahuan Wei, Jiayu Li, Yanfeng Sun, Jia Cheng, Zhanyi Li, Zhiyuan Yao, Xu Wang, Xinyi Li, Danxu Cui, Xuan Cao, Jun Li, Cheng Zheng, Jianbin Sarvaiya, Anjali Prajapati, Kalpesh Patra, Ratnadeep Barik, Pragnesh Rathod, Chaitanya Upla, Kishor Raja, Kiran Ramachandra, Raghavendra Busch, Christoph Computer Vision Lab Eth Zurich Switzerland Shanghai Jiao Tong University China University of Würzburg Germany Xidian University China Mais&cripac Institute of Automation Chinese Academy of Sciences China School of Artificial Intelligence University of Chinese Academy of Sciences China University of Science and Technology of China China Beijing Institute of Technology China School of Information Science and Technology ShanghaiTech University China School of Information and Communication Engineering University of Electronic Science and Technology of China China China Lotte Data Communication Company Seoul Korea Republic of Graduate Institute of Electronics Engineering National Taiwan University Taiwan Department of Electrical Engineering National Taiwan University Taiwan Graduate Institute of Communication Engineering National Taiwan University Taiwan ServiceNow United States MegaStudyEdu Korea Republic of Computer Vision Lab Caidas University of Würzburg Germany University of Sindh Pakistan Iklab Inc. Tech University of Korea Siheung-Si Korea Republic of Micro-Nano Electronics Department Shanghai Jiao Tong University China Electrical Engineering Department Faculty of Engineering Assiut University Egypt Hefei University of Technology China Beijing Jiaotong University China South China University of Technology Guangdong Guangzhou China Sardar Vallabhbhai National Institute of Technology India Norwegian University of Science and Technology Norway

ISBN: (纸本)9798350302493

This paper reviews the NTIRE 2023 challenge on image super-resolution (×4), focusing on the proposed solutions and results. The task of image super-resolution (SR) is to generate a high-resolution (HR) output from a corresponding low-resolution (LR) input by leveraging prior information from paired LR-HR images. The aim of the challenge is to obtain a network design/solution capable to produce high-quality results with the best performance (e.g., PSNR). We want to explore how high performance we can achieve regardless of computational cost (e.g., model size and FLOPs) and data. The track of the challenge was to measure the restored HR images with the ground truth HR images on DIV2K testing dataset. The ranking of the teams is determined directly by the PSNR value. The challenge has attracted 192 registered participants, where 15 teams made valid submissions. They achieve state-of-the-art performance in single image super-resolution. © 2023 IEEE.

关键词： Statistical tests

来源：评论

学校读者我要写书评

暂无评论

Designing Resilient Network Architectures for Disaster Recovery

Designing Resilient Network Architectures for Disaster Recov...

引用

Innovations in High Speed Communication and Signal Processing (IHCSP), International Conference on

作者： Parikshit N. Mahalle A. Amudha Anubhav Bhalla Jagtej Singh Janakiraman V Srinivasa Rao K S Department of Artificial Intelligence & Data Science Vishwakarma Institute of Technology Pune India Department of Electrical and Electronics Engineering Karpagam Academy of Higher Education Coimbatore Centre of Research Impact and Outcome Chitkara University Rajpura Punjab India Chitkara Centre for Research and Development Chitkara University Himachal Pradesh India Department of Mech Prince Shri Venkateshwara Padmavathy Engineering College Chennai India Adjunct Faculty School of Mangement-PG Management JAIN Deemed to be University Bangalore Karnataka India

ISBN: (数字)9798350368949

ISBN: (纸本)9798350368956

Digital infrastructure is always changing, therefore secure network designs are essential to keep things running amid natural catastrophes. The Dynamic Resilience Framework (DRF) uses ARO, DDI, and ERA to strengthen networks. Many prominent methods are used to compare the proposed framework against new ones. Hierarchical network design, SDN, distributed cloud architecture, and active-active redundancy are examples. The experiment simulates disaster situations to examine each network architecture's downtime, data recovery, and resource use. DDI scores reveal how well components interact; ARO scores show how repetition varies with difficulty. ERA scores measure resource efficiency and how to scale them with demand. You may analyze qualitative and quantitative statistics, user comments, and qualitative assessments. DRF outperforms other methods in ARO, DDI, and ERA. Statistics show that these performance variances are considerable, proving that the proposed structure is strong and adaptive. This research provides a comprehensive catastrophic recovery strategy, adding to the resilient network architecture discussion. The results will help companies determine the best ways to protect their digital assets against unexpected challenges.

关键词： Technological innovation Disasters Scalability Redundancy Network architecture Signal processing Resource management Software defined networking Resilience Testing

来源：评论

学校读者我要写书评

暂无评论

LogisticsVISTA: 3D Terminal Delivery Services With UAVs, UGVs and USVs Based on Foundation Models and Scenarios engineering

LogisticsVISTA: 3D Terminal Delivery Services With UAVs, UGV...

引用

2024 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI 2024

作者： Tian, Yonglin Lin, Fei Zhang, Xinyuan Ge, Jingwei Wang, Yutong Dai, Xingyuan Lv, Yisheng Wang, Fei-Yue University of Chinese Academy of Sciences Beijing100049 China Tsinghua University Department of Automation Beijing100084 China Institute of Automation Chinese Academy of Sciences State Key Laboratory of Multimodal Artificial Intelligence Systems State Key Laboratory for Management and Control of Complex Systems Beijing100190 China Macau University of Science and Technology Faculty of Innovation Engineering Department of Engineering Science 999078 China

Intelligent logistics with autonomous vehicles is promising to improve the efficiency of rapidly expanding delivery systems. However, complex environments, diverse packages, and irregular human interaction have imposed challenges to the application of unmanned delivery vehicles. Combining various forms of vehicles and developing knowledge-driven delivery services have been the key to further facilitating the upgrading of logistics systems. In this paper, we propose an integrated framework named LogisticsVISTA (Logistics on Vehicles with Intelligent Systems for Transport Automation) comprised of UAVs (Unmanned Aerial Vehicles), UGVs (Unmanned Ground Vehicles), and USVs (Unmanned Surface Vehicles) to provide timely and flexible 3D delivery services. In LogisticsVISTA, we design Large language models (LLMs) and large multimodal models (LMMs) for task planning, active perception, autonomous decision, and human-machine interaction. Besides, virtual-real integrated parallel scenarios are constructed to provide the playing ground for embodied agents efficiently and flexibly. © 2024 IEEE.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Ethical Considerations in Explainable AI: Balancing Transparency and User Privacy in English Language-based Virtual Assistants

Ethical Considerations in Explainable AI: Balancing Transpar...

引用

Intelligent Communication Technologies and Virtual Mobile Networks (ICICV), International Conference on

作者： Franciskus Antonius Alijoyo S. Sandya Sneha Sri Purnachandra Rao Alapati Dilyorjon Yuldashev Ponni Valavan M Center for Risk Management and Sustainability Indonesia School of Business and Information Technology STMIK LIKMI Bandung Indonesia Krishna University Andhra Pradesh Prasad V Potluri Siddhartha Institute of Technology Kanuru Vijayawada Andhra Pradesh India Automatic Control and Computer Engineering Department Turin Polytechnic University in Tashkent Tashkent Uzbekistan Department of Artificial Intelligence and Data Science K.Ramakrishnan college of Engineering Trichy Tamil Nadu India

English Language-Based Virtual Assistants (ELB-VAs) are AI-powered systems designed to comprehend and respond to user queries in the English language, exemplified by virtual assistants like Siri or Alexa. The need for balancing transparency and user privacy in ELB-VAs is paramount due to their pervasive integration into daily life. Ensuring transparency imbues user trust, while safeguarding privacy addresses ethical concerns associated with personal data. Existing methods involve clear privacy policies, user-controlled data sharing settings, and encryption. However, drawbacks include user confusion and potential biases. To address these limitations, this study proposes a novel approach. Methodologically, it integrates pre-processing techniques such as lowercasing and tokenization, coupled with a Natural Language Understanding model. This model undergoes intent and entity recognition training, enhancing accuracy, and incorporates privacy-aware response generation, ensuring informative yet privacy-conscious interactions. The implementation of the study's results is carried out using Python tools, showcasing improved metrics and response times. This approach contributes to a more transparent and privacy-respecting user experience, aligning with evolving ethical norms and setting the stage for advancements in ELB-VA technology. This comprehensive exploration bridges existing gaps, emphasizing the ethical imperative of user-centric and privacy-aware AI interactions in ELB- VAs. The proposed NLU model exhibits a substantial increase in accuracy compared to other methods, with an impressive accuracy value of 99.1 % • On average, it outperforms the Random Forest and Decision Tree models by 15.7 percentage points, highlighting its superior predictive capabilities in the evaluated task. This comprehensive exploration aligns with evolving ethical norms and establishes a foundation for future advancements in ELB-VA technology.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Digital Twin-enabled Intelligent DDoS Detection Mechanism for Autonomous Core Networks

arXiv

引用

arXiv 2023年

作者： Yigit, Yagmur Bal, Bahadir Karameseoglu, Aytac Duong, Trung Q. Canberk, Berk The Department of Computer Engineering Faculty of Computer and Informatics Istanbul Technical University Turkey The Microsoft Denmark The BTS Group Istanbul Turkey School of Electronics Electrical Engineering and Computer Science Queen’s University Belfast United Kingdom The School of Engineering and Built Environment Edinburgh Napier University United Kingdom The Department of Artificial Intelligence and Data Engineering Istanbul Technical University Turkey

Existing distributed denial of service attack (DDoS) solutions cannot handle highly aggregated data rates;thus, they are unsuitable for Internet service provider (ISP) core networks. This paper proposes a digital twin-enabled intelligent DDoS detection mechanism using an online learning method for autonomous systems. Our contributions are three-fold: we first design a DDoS detection architecture based on the digital twin for ISP core networks. We implemented a Yet Another Next Generation (YANG) model and an automated feature selection (AutoFS) module to handle core network data. We used an online learning approach to update the model instantly and efficiently, improve the learning model quickly, and ensure accurate predictions. Finally, we reveal that our proposed solution successfully detects DDoS attacks and updates the feature selection method and learning model with a true classification rate of ninety-seven percent. Our proposed solution can estimate the attack within approximately fifteen minutes after the DDoS attack starts. © 2023, CC BY.

关键词： Denial-of-service attack

来源：评论

学校读者我要写书评

暂无评论

Study of Structural Optimization on Mg-1ca Interference Screws

SSRN

引用

SSRN 2024年

作者： Shen, Zhenquan Zhou, Xiaochen Bian, Dong Zhao, Ming Li, Yafei Zheng, Yufeng Faculty of Artificial Intelligence in Education Central China Normal University Wuhan430079 China School of Materials Science and Engineering Peking University Beijing100871 China Department of Orthopedics Guangdong Provincial People’s Hospital Guangdong Academy of Medical Sciences Guangzhou510080 China Schlumberger Technology Corporation HoustonTX77054 United States School of Materials Science and Engineering Zhengzhou University Zhengzhou450001 China Department of Materials Science and Engineering China

Biodegradable magnesium alloys have attracted great interest as the matrix materials for the next generation of absorbable metallic interference screws. It is necessary to optimize the structure so that screws can work better in vivo since their degradation is highly related to structural design. In this paper, Optimal Latin Hypercube Sampling technique (OLHS) is used to sample in the design space. The pull-out strengths of sample points as optimization objective are calculated through a continuum damage model (CDM) implemented into a finite element analysis (FEA). Response surface model (RSM) and Kriging model are respectively used as surrogate model for this optimization problem to compare. Both of the surrogate models are optimized based on a sequential quadratic programming (SQP) strategy. It is found that the result of Kriging model is acceptable after FEA verification. Relationship between parameters, rationality and limitation of this method are discussed. In the end, a final design and some advice about interference screw design are provided. © 2024, The Authors. All rights reserved.

关键词： Finite element method

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：