检索结果-内蒙古大学图书馆

38th Conference on Neural information Processing Systems, NeurIPS 2024

作者： Zhao, Dong Zang, Qi Wang, Shuang Sebe, Nicu Zhong, Zhun School of Artificial Intelligence Xidian University Shaanxi China Department of Information Engineering and Computer Science University of Trento Italy School of Computer Science and Information Engineering Hefei University of Technology China School of Computer Science University of Nottingham NottinghamNG8 1BB United Kingdom

Presently, pseudo-labeling stands as a prevailing approach in cross-domain semantic segmentation, enhancing model efficacy by training with pixels assigned with reliable pseudo-labels. However, we identify two key limitations within this paradigm: (1) under relatively severe domain shifts, most selected reliable pixels appear speckled and remain noisy. (2) when dealing with wild data, some pixels belonging to the open-set class may exhibit high confidence and also appear speckled. These two points make it difficult for the pixel-level selection mechanism to identify and correct these speckled close- and open-set noises. As a result, error accumulation is continuously introduced into subsequent self-training, leading to inefficiencies in pseudo-labeling. To address these limitations, we propose a novel method called Semantic Connectivity-driven Pseudo-labeling (SeCo). SeCo formulates pseudo-labels at the connectivity level, which makes it easier to locate and correct closed and open set noise. Specifically, SeCo comprises two key components: Pixel Semantic Aggregation (PSA) and Semantic Connectivity Correction (SCC). Initially, PSA categorizes semantics into "stuff" and "things" categories and aggregates speckled pseudo-labels into semantic connectivity through efficient interaction with the Segment Anything Model (SAM). This enables us not only to obtain accurate boundaries but also simplifies noise localization. Subsequently, SCC introduces a simple connectivity classification task, which enables us to locate and correct connectivity noise with the guidance of loss distribution. Extensive experiments demonstrate that SeCo can be flexibly applied to various cross-domain semantic segmentation tasks, i.e. domain generalization and domain adaptation, even including source-free, and black-box domain adaptation, significantly improving the performance of existing state-of-the-art methods. The code is available at https://***/DZhaoXd/SeCo. © 2024 Neural information

关键词：

来源：评论

学校读者我要写书评

暂无评论

Occupancy estimation in smart buildings using predictive modeling in imbalanced domains

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第8期14卷 10917-10929页

作者： Guo, Jiaxun Amayri, Manar Najar, Fatma Fan, Wentao Bouguila, Nizar Concordia Institute for Information Systems Engineering Concordia University MontrealQCH3G1T7 Canada Grenoble Institute of Technology Grenoble France The Department of Computer Science and Technology Huaqiao University Xiamen China

This paper introduces a novel approach for occupancy estimation in smart buildings. In particular, we focus on the challenging yet common situation where the amount of training data is small and imbalanced (i.e. the classes are not approximately equally represented). Our model is based on two parts namely predictive modelling, performed via the inverted Dirichlet mixture (IDMM), and an over-sampling approach that we propose. The first part, in which the main goal is to tackle the small training data problem, concerns the calculation of the predictive distribution of the IDMM by marginalizing over its parameters, with their posterior distributions, which are estimated by a Bayesian variational inference approach that we develop. Based on over-sampling, the second part can be viewed as a complement to tackling the imbalanced domains problem. Extensive experiments and simulations involving synthetic data as well as real data extracted from smart buildings sensors show the merits of our statistical framework. © 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Intelligent buildings

来源：评论

学校读者我要写书评

暂无评论

Self-Learning Based Dependable Offloading Optimization in Semi-Trusted Vehicular Edge Computing and Networks

引用

IEEE Transactions on Vehicular technology 2025年第5期74卷 8141-8157页

作者： Li, Xuehan Jing, Tao Li, Ruinian Wang, Xiaoxuan Yan, Yu Fan, Xin Huo, Yan Yu, Fei Richard Beijing Jiaotong University School of Electronic and Information Engineering National Engineering Research Center of Advanced Network Technologies Beijing100044 China Bowling Green State University Department of CS Bowling GreenOH43403 United States Beijing Forestry University School of Information Science and Technology Beijing100083 China Carleton University Department of Systems and Computer Engineering OttawaONK1S 5B6 Canada

As vehicular edge computing and networks (VECONs) emerge, Internet of vehicles (IoVs) is gaining significant advantages. To provide users with dependable services, VECONs need to offer efficient and secure task offloading services. Unfortunately, little effort has been put into addressing security issues such as the risk of privacy exposure when offloading data. Modeling and optimizing offloading dependability are affected by these issues. To address these concerns, a dependable three-layer offloading framework involving blockchain technology is presented in this paper, with fully homomorphic encryption (FHE) algorithm incorporated in order to mitigate privacy exposure risks. The framework divides tasks reasonably and introduces a novel measure of offloading dependability called cost from energy consumption, delay, and privacy exposure risk (cEDP) for constructing dependable offloading optimization. Additionally, a self-learning based dependable privacy preservation offloading (SDPPO) algorithm is presented that is capable of solving the optimal offloading problem in a decentralized manner. Simulation results show that the SDPPO algorithm outperforms its counterparts in terms of faster convergence speed, better cEDP convergence results, and improved scalability and tunability characteristics. © 1967-2012 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

A Temporal-Spatial Graph Network With a Learnable Adjacency Matrix for Appliance-Level Electricity Consumption Prediction

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第4期6卷 989-1002页

作者： Li, Dandan Xia, Jiaxing Li, Jiangfeng Xiao, Changjiang Stankovic, Vladimir Stankovic, Lina Shi, Qingjiang Tongji University School of Computer Science and Technology Shanghai201804 China China University of Mining and Technology Artificial Intelligence Research Institute Jiangsu Xuzhou221008 China Tongji University College of Surveying and Geo-Information Shanghai200092 China University of Strathclyde Department of Electronic and Electrical Engineering GlasgowG1 1XJ United Kingdom

Predicting the electricity consumption of individual appliances, known as appliance-level energy consumption (ALEC) prediction, is essential for effective energy management and conservation. Despite its importance, research in this area is limited and faces several challenges: 1) the correlation between the usage of different appliances has rarely been considered for ALEC prediction;2) a learnable strategy for obtaining the optimal correlation between different appliance behaviors is lacking;and 3) it is difficult to accurately quantify the usage relationship among different appliances. To address these issues, we propose a graph-based temporal-spatial network that employs a learnable adjacency matrix for appliance-level load prediction in this work. The network comprises a temporal graph convolutional network (TGCN) and a learnable adjacency matrix that enables us to utilize correlations between appliances and quantify their relationships. To validate our approach, we compared our model with six others: a TGCN model with a fixed adjacency matrix where all elements are set to 0;a TGCN model with a fixed adjacency matrix where all elements are set to 0.5, except for the diagonal;a TGCN model with a randomly generated adjacency matrix, except for the diagonal;an Aug-LSTM model;a model with ResNetPlus architecture;and a feed-forward deep neural network. Five houses in four datasets: AMPDs, REFIT, UK-DALE, and SC-EDNRR are utilized. The metrics used in this study include root mean square error, explained variance score, mean absolute error, F-norm and coefficient of determination. Our experiments have validated the accuracy and practicality of our proposed approach across different datasets. © 2024 IEEE.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Ontology-Based Crime News Semantic Retrieval System

引用

computers, Materials & Continua 2023年第10期77卷 601-614页

作者： Fiaz Majeed Afzaal Ahmad Muhammad Awais Hassan Muhammad Shafiq Jin-Ghoo Choi Habib Hamam Department of Information Technology University of GujratGujrat50700Pakistan Department of Computer Science University of Engineering&TechnologyLahore54890Pakistan Department of Information and Communication Engineering Yeungnam UniversityGyeongsan38541Korea Faculty of Engineering Uni de MonctonMonctonNB E1A3E9Canada International Institute of Technology and Management Commune d’AkandaBPLibreville1989Gabon School of Electrical Engineering Department of Electrical and Electronic Engineering ScienceUniversity of JohannesburgJohannesburg2006South Africa Spectrum of Knowledge Production&Skills Development Sfax3027Tunisia

Every day,the media reports tons of crimes that are considered by a large number of users and accumulate on a regular *** news exists on the Internet in unstructured formats such as books,websites,documents,and *** such homogeneous data,it is very challenging to extract relevant information which is a time-consuming and critical task for the public and law enforcement ***-based information Retrieval(IR)systems rely on statistics to retrieve results,making it difficult to obtain relevant *** are unable to understandthe user’s query and thus facewordmismatchesdue to context changes andthe inevitable semanticsof a given ***,such datasets need to be organized in a structured configuration,with the goal of efficiently manipulating the data while respecting the semantics of the *** ontological semantic IR systemis needed that can find the right investigative information and find important clues to solve criminal *** semantic system retrieves information in view of the similarity of the semantics among indexed data and user *** this paper,we develop anontology-based semantic IRsystemthat leverages the latest semantic technologies including resource description framework(RDF),semantic protocol and RDF query language(SPARQL),semantic web rule language(SWRL),and web ontology language(OWL).We have conducted two *** the first experiment,we implemented a keyword-based textual IR systemusing Apache *** the second experiment,we implemented a semantic systemthat uses ontology to store the data and retrieve precise results with high accuracy using SPARQL *** keyword-based system has filtered results with 51%accuracy,while the semantic system has filtered results with 95%accuracy,leading to significant improvements in the field and opening up new horizons for researchers.

关键词： Web 3.0 crime ontology semantic web knowledge representation

来源：评论

学校读者我要写书评

暂无评论

Multidomain Object Detection Framework Using Feature Domain Knowledge Distillation

引用

IEEE Transactions on Cybernetics 2024年第8期54卷 4643-4651页

作者： Jaw, Da-Wei Huang, Shih-Chia Lu, Zhi-Hui Fung, Benjamin C. M. Kuo, Sy-Yen National Taiwan University Department of Electrical Engineering Taipei106 Taiwan National Taipei University of Technology Department of Electronic Engineering Taipei106 Taiwan Fudan University School of Computer Science The Engineering Research Center of Cyber Security Auditing and Monitoring Ministry of Education Shanghai200433 China McGill University School of Information Studies MontrealQCH3A 1X1 Canada

Object detection techniques have been widely studied, utilized in various works, and have exhibited robust performance on images with sufficient luminance. However, these approaches typically struggle to extract valuable features from low-luminance images, which often exhibit blurriness and dim appearence, leading to detection failures. To overcome this issue, we introduce an innovative unsupervised feature domain knowledge distillation (KD) framework. The proposed framework enhances the generalization capability of neural networks across both low- and high-luminance domains without incurring additional computational costs during testing. This improvement is made possible through the integration of generative adversarial networks and our proposed unsupervised KD process. Furthermore, we introduce a region-based multiscale discriminator designed to discern feature domain discrepancies at the object level rather than from the global context. This bolsters the joint learning process of object detection and feature domain distillation tasks. Both qualitative and quantitative assessments shown that the proposed method, empowered by the region-based multiscale discriminator and the unsupervised feature domain distillation process, can effectively extract beneficial features from low-luminance images, outperforming other state-of-the-art approaches in both low- and sufficient-luminance domains. © 2013 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Industrial Automation information Analogy for Smart Grid Security

引用

computers, Materials & Continua 2022年第5期71卷 3985-3999页

作者： Muhammad Asif Ishfaq Ali Shahbaz Ahmad Azeem Irshad Akber Abid Gardezi Fawaz Alassery Habib Hamam Muhammad Shafiq Department of Computer Science National Textile UniversityFaisalabad37610PunjabPakistan Department of Computer Science and Software Engineering International Islamic UniversityIslamabadPakistan Department of Computer Science COMSATS University IslamabadIslamabadPakistan Department of Computer Engineering College of Computers and Information TechnologyTaif UniversityTaifSaudi Arabia Faculty of Engineering Moncton UniversityNBE1A3E9Canada Department of Information and Communication Engineering Yeungnam UniversityGyeongsan38541Korea

Industrial automation or assembly automation is a strictly monitored environment,in which changes occur at a good *** are many types of entities in the focusing environment,and the data generated by these devices is *** addition,because the robustness is achieved by sensing redundant data,the data becomes *** data generating device,whether it is a sensing device or a physical device,streams the data to a higher-level deception device for calculation,so that it can be driven and configured according to the updated *** the emergence of the Industry 4.0 concept that includes a variety of automation technologies,various data is generated through numerous ***,the data generated for industrial automation requires unique information Architecture(IA).IA should be able to satisfy hard real-time constraints to spontaneously change the environment and the instantaneous configuration of all *** understand its applicability,we used an example smart grid *** smart grid system needs an IA to fulfill the communication requirements to report the hard real-time changes in the power immediately following the *** addition,in a smart grid system,it needs to report changes on either side of the system,i.e.,consumers and suppliers configure and reconfigure the system according to the *** this article,we propose an analogy of a physical phenomenon.A point charge is used as a data generating device,the streamline of electric flux is used as a data flow,and the charge distribution on a closed surface is used as a ***,the intensity changes are used in the physical process,e.g.,the smart *** analogy is explained by metaphors,and the structural mapping framework is used for its theoretical *** proposed analogy provides a theoretical basis for the development of such information architectures that can represent data flows,definition changes(deterministic and non-deterministic),events,and i

关键词： Industry 4.0 information architecture industrial automation

来源：评论

学校读者我要写书评

暂无评论

Automated Grading of Breast Cancer Histopathology Images Using Multilayered Autoencoder

引用

computers, Materials & Continua 2022年第5期71卷 3407-3423页

作者： Shakra Mehak M.Usman Ashraf Rabia Zafar Ahmed M.Alghamdi Ahmed S.Alfakeeh Fawaz Alassery Habib Hamam Muhammad Shafiq Knowledge Unit of System and Technology University of Management&TechnologySialkot Campus51310Pakistan Department of Computer Science University of Management and TechnologySialkot Campus51310Pakistan Faculty of Computing and Technology University of Engineering&TechnologyNarowal51610Pakistan College of Computer Science and Engineering University of Jeddah21493Saudi Arabia Faculty of Computing and Information Technology King Abdulaziz UniversityJeddah21589Saudi Arabia Department of Computer Engineering College of Computers and Information TechnologyTaif UniversityTaifSaudi Arabia Faculty of Engineering Moncton UniversityNBE1A3E9Canada Department of Information and Communication Engineering Yeungnam UniversityGyeongsan38541Korea

Breast cancer(BC)is the most widely recognized cancer in women *** 2018,627,000 women had died of breast cancer(World Health Organization Report 2018).To diagnose BC,the evaluation of tumours is achieved by analysis of histological *** present,the Nottingham Bloom Richardson framework is the least expensive approach used to grade BC *** contemplate three elements,*** count,*** formation,and *** atypia,which is a laborious process that witness’s variations in expert’s ***,some algorithms have been proposed for the detection of mitotic cells,but nuclear atypia in breast cancer histopathology has not received much *** atypia analysis is performed not only to grade BC but also to provide critical information in the discrimination of normal breast,non-invasive breast(usual ductal hyperplasia,atypical ductal hyperplasia)and pre-invasive breast(ductal carcinoma in situ)and invasive breast *** proposed a deep-stacked multi-layer autoencoder ensemble with a softmax layer for the feature extraction and classification *** classification results show the value of the multilayer autoencoder model in the evaluation of nuclear *** proposed method has indicated promising results,making them more fit in breast cancer grading.

关键词： Breast cancer nuclear atypia autoencoder

来源：评论

学校读者我要写书评

暂无评论

An Edge Computing System for Fast Image Recognition Based on Convolutional Neural Network and Petri Net Model

引用

Multimedia Tools and Applications 2024年第5期83卷 12849-12873页

作者： Yang, Cheng-Ying Lin, Yi-Nan Wang, Sheng-Kuan Shen, Victor R.L. Tung, Yi-Chih Lin, Jia-Fu Department of Computer Science University of Taipei 1 Ai-Kao W. Rd. Taipei 100 Taipei Taiwan Department of Electronic Engineering Ming Chi University of Technology 84 Gongzhuan Rd. Taishan District New Taipei City 243 New Taipei Taiwan Department of Electrical Engineering Ming Chi University of Technology 84 Gongzhuan Rd. Taishan District New Taipei City 243 New Taipei Taiwan Department of Computer Science and Information Engineering National Taipei University 151 University Rd. Sanxia District New Taipei City 237 New Taipei Taiwan Department of Information Management Chaoyang University of Technology Taichung Taiwan

As a computer system can precisely detect some target objects, many application scenarios will be developed. In this study, the customized object recognition model was trained by using a software tool, You Only Look Once (YOLO), and burned into one hardware component, Maix Bit, for an image edge computing system, so as to fast recognize humans or objects in the images. In the beginning of system development, a Petri net (PN) model was established to detect all possible abnormal processes and to verify the feasibility and completeness of an edge computing system by using Petri net software tool, Workflow Petri net Designer (WoPeD). Compared with many other development boards, Maix Bit is smaller, more flexible, more economical, and more adaptable to those needs of the edge side scenarios. When equipped with an Artificial Intelligence (AI) chip, Kendryte K210, it is suitable for object recognition due to its superior power consumption, film frames, and processors. YOLO, mainly characterized by its fast speed in using the convolutional neural network (CNN) to recognize objects, can be used to predict the positions and types of multiple objects concurrently. Also, CNN can be used to detect and recognize targets from end to end promptly, when configured to train an image recognition model. Finally, the experimental results have shown that the promising mean Average Precision (mAP), 78.6%, was obtained, which outperforms other existing systems. © 2023, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Effect of Intelligent Multi-Association in Civil Aircraft-Augmented SAGIN

引用

IEEE Transactions on Cognitive Communications and Networking 2023年第1期9卷 223-238页

作者： Chen, Qian Meng, Weixiao Han, Shuai Li, Cheng Chen, Hsiao Hwa Harbin Institute of Technology School of Electronics and Information Engineering Harbin150001 China Memorial University Department of Electrical and Computer Engineering St. John'sA1B 3X5 Canada National Cheng Kung University Department of Engineering Science Tainan70101 Taiwan

Although 5G networks have gradually commercialized, many scenarios like emergency areas and remote regions still exist with vast communication problems. In this paper, we provide a detailed analysis of civil-aircraft augmented space-air-ground integrated networks (CAA-SAGIN), where the user's association strategy can be determined intelligently. First, we discuss the influence of the spatial distribution of civil aircrafts (CAs) and satellites on the Rician factor. Then, we provide a novel and simple expression of moment-generating functions related to small-fading variables. With massive visible access points (APs), we combine two user-centric connectivity methods, i.e., multiple resource blocks in a single AP (MRB-SA) and collaborative connectivity in multiple APs (CC-MA), and each segment's ergodic capacity and coverage probability are investigated respectively. The numerical results demonstrate that CA networks can provide similar coverage and capacity performance as satellite networks in the given range. The advantages of introducing CAs to the satellite-terrestrial integrated networks are significant under the combined connectivity methods. This paper quantitatively proves CA networks' benefits in capacity and coverage and lays a theoretical foundation for CAA-SAGIN. © 2015 IEEE.

关键词： Satellites

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：