检索结果-内蒙古大学图书馆

Asymmetric network pseudo labels mutual refinement for unsupervised domain adaptation person re-identification

Multimedia Tools and Applications 2024年第40期83卷 88091-88111页

作者： Yun, Xiao Chen, Jiawang Zhang, Xiaoguang Dong, Kaiwen Li, Song Sun, Yanjing School of Information and Control Engineering China University of Mining and Technology Xuzhou221116 China Xi’an Key Laboratory of Heterogeneous Network Convergence Communication Xi’an710054 China

In the task of unsupervised domain adaptation person re-identification, the traditional symmetric dual-branch network only generates one single feature, which ignores the difference and complementarity of the network and is prone to produce wrong cluster results. To solve this problem, the asymmetric network pseudo labels mutual refinement (ANPR) algorithm is proposed to design an asymmetric network (ASNet), containing diverse features for mutual supervision. One branch of ASNet employs convolutional operations with constrained receptive fields to extract local information, and the other branch uses the contextual transformer block to capture global features. Secondly, this paper constructs the pseudo labels mutual refinement (PMR) module, which generates two sets of clustering results using global and local potential features of unlabeled samples. PMR suppresses the inconsistent clustering results and retains the consistent ones, which gradually improves the quality of pseudo labels for dual-brach mutual supervision. In addition, the feature space loss (FSL) function is designed to focus on the relative distance of samples in the feature layer without relying on the pseudo labels in the task layer, which effectively avoids the influence of noisy pseudo labels in the optimization process. The proposed method is evaluated on three popular datasets, and extensive experimental results demonstrate its effectiveness. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Coherent range-dependent map-drift algorithm for improving SAR motion compensation

引用

Journal of Systems engineering and Electronics 2023年第1期34卷 47-55页

作者： ZHANG Man WANG Guanyong WEI Feiming JIN Xue School of Electronic and Communication Engineering Guangzhou UniversityGuangzhou 510006China Beijing Institute of Radio Measurement Beijing 100854China School of Electronic Information and Electrical Engineering Shanghai Jiao Tong UniversityShanghai 200030China No.802 Research Institute of Shanghai Academy of Spaceflight Technology Shanghai 200050China Beijing Institute of Tracking and Communication Technology Beijing 100094China

Synthetic aperture radar(SAR) is usually sensitive to trajectory deviations that cause serious motion error in the recorded data. In this paper, a coherent range-dependent mapdrift(CRDMD) algorithm is developed to accommodate the range-variant motion errors. By utilizing the algorithm as an estimate core, robust motion compensation strategy is proposed for unmanned aerial vehicle(UAV) SAR imagery. CRDMD outperforms the conventional map-drift algorithms in both accuracy and efficiency. Real data experiments show that the proposed approach is appropriate for precise motion compensation for UAV SAR.

关键词： synthetic aperture radar(SAR) motion compensation(MOCO) coherent map-drift(CMD) range-variant

来源：评论

学校读者我要写书评

暂无评论

Multi-Object Detection and Classification in Construction Sites Based on YOLOv5 5

Multi-Object Detection and Classification in Construction Si...

引用

5th International Conference on Video, Signal and Image Processing, VSIP 2023

作者： Liang, Xintao Jiang, Tianqi Fu, Qiang Wang, Qingyan School Of Measurement And Control Technology And Telecommunications Engineering Harbin University Of Science And Technology Harbin150080 China

ISBN: (纸本)9798400709272

A thorough analysis of the framework structure of the YOLO algorithm is conducted, and based on the YOLOv5 algorithm, rapid detection and classification of extracted features are implemented. Addressing the issue of multi-object detection and classification in engineering sites, this study utilizes the YOLOv5 algorithm for object detection in engineering scenarios, constructs the ResNet50 network, and achieves training and recognition of categories for engineering vehicles, whether washed or unwashed. Additionally, the YOLOv5 algorithm is employed to detect whether construction personnel are wearing safety helmets and reflective clothing, enabling fast and accurate entity detection and classification at construction sites. © 2023 Owner/Author.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Impact Damage Identification of Aluminum Alloy Reinforced Plate Based on GWO-ELM Algorithm

引用

Structural Durability & Health Monitoring 2023年第6期17卷 485-500页

作者： Wei Li Benjian Zou Yuxiang Luo Ning Yang Faye Zhang Mingshun Jiang Lei Jia School of Control Science and Engineering Shandong UniversityJinan250061China Communication Technology Division Shandong Institute of Space Electronic TechnologyYantai264000China

As a critical structure of aerospace equipment,aluminum alloy stiffened plate will influence the stability of spacecraft in orbit and the normal operation of the *** this study,a GWO-ELM algorithm-based impact damage identification method is proposed for aluminum alloy stiffened panels to monitor and evaluate the damage condition of such stiffened panels of ***,together with numerical simulation,the experimental simulation to obtain the damage acoustic emission signals of aluminum alloy reinforced panels is performed,to establish the damage ***,the amplitude-frequency characteristics of impact damage signals are extracted and put into an extreme learning machine(ELM)model to identify the impact location and damage degree,and the Gray Wolf Optimization(GWO)algorithm is employed to update the weight parameters of the ***,experiments are conducted on the irregular aluminum alloy stiffened plate with the size of 2200 mm×500 mm×10 mm,the identification accuracy of impact position and damage degree is 98.90% and 99.55% in 68 test areas,*** experiments with ELM and backpropagation neural networks(BPNN)demonstrate that the impact damage identification of aluminum alloy stiffened plate based on GWO-ELM algorithm can serve as an effective way to monitor spacecraft structural damage.

关键词： GWO-ELM aluminum alloy stiffened plate damage identification amplitude-frequency characteristic

来源：评论

学校读者我要写书评

暂无评论

Failure Mode Analysis and Infrared Image Recognition of Crystalline Silicon PV Module 23

Failure Mode Analysis and Infrared Image Recognition of Crys...

引用

Proceedings of the 2023 5th International Conference on Video, Signal and Image Processing

作者： Ziyi Li Lijie Wang Jianbo Sun Xinyu Lu Xianwen Zhu School of Measurement and Control Technology and Communication Engineering Harbin University of Science and Technology China

ISBN: (纸本)9798400709272

In this paper, the fault analysis of crystalline silicon photovoltaic modules is studied. The Failure Mode and Effect Analysis (FMEA) and Fault Tree Analysis (FTA) methods are used to analyze the failure mode and its causes. Based on the results of fault analysis, the infrared image analysis and recognition methods are studied for the two fault modes of infrared hot spot and component shedding of Photovoltaic(PV) modules. The YOLOv5s image recognition method based on clustering improvement and feature enhancement is discussed. The experimental results show that the YOLOv5s image recognition algorithm with clustering improvement and feature enhancement improves the training effect of the model by using the EIOU loss function to adaptively adjust the confidence loss balance coefficient, and the detection speed (Frame Per Second, FPS) can reach 42.37 FPS; by adding InRe feature enhancement modules before each detection layer, the extraction ability of target features is improved. The mean Average Precision (mAP) of hot spot and component shedding are 94.85 % and 90.67 %, respectively, which can fully meet the needs of UAV automatic inspection.

关键词： FMEA FTA Image recognition Infrared image PV module

来源：评论

学校读者我要写书评

暂无评论

Multi-scale network for single image deblurring based on ensemble learning module

引用

Multimedia Tools and Applications 2025年第11期84卷 9045-9064页

作者： Wu, WenBo Pan, Yun Su, Na Wang, JingTao Wu, ShaoChuan Xu, ZeKun Yu, YouJian Liu, YaPeng State Key Laboratory of Media Convergence and Communication Communication University of China Beijing100024 China School of Computer and Cyberspace Security Communication University of China Beijing100024 China School of Electronics and Information Engineering Harbin Institute of Technology Harbin150080 China College of Computer and Control Engineering Northeast Forestry University Harbin150040 China School of Computer and Information Engineering Tianjin Chengjian University Tianjin300384 China School of Computer Science and Technology Anhui University Hefei230601 China

In this paper, we have identified two primary issues with current multi-scale image deblurring methods. On the one hand, the blurring scale is ignored. On the other hand, the context information of images is not fully utilized, and the spatial detail information of images is lost. To better solve the above problems, on top of that, this paper proposes a multi-scale network for single image deblurring. The designed network contains two parts: 1) Ensemble Learning Model: The Weak Learning Model in this module connects multiple weak learners in series to increase network depth. Moreover, through multi-scale feature ensemble module, the output features and intermediate features of different weak learner are integrated, and finally strong learner is used to enhance the features with semantic information at different scales;2) Spatial Feature Enhancement Module: This module can preserve spatial detail information of images and compensate for the disadvantages of losing image spatial details due to continuous downsampling in the Ensemble Learning Model structure. Our proposed deblurring model has been extensively evaluated on several benchmark datasets, and the model achieves superior performance compared to the most advanced deblurring methods. In particular, the peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) values are 33.57 dB and 0.965 on the GoPro dataset, respectively, with PSNR and SSIM improving by 0.49 dB and 0.002 over NAFNet, and 0.65 dB and 0.004 over Restormer. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Torque Ripple Investigation on Line-start Permanent Magnet Synchronous Motors

Torque Ripple Investigation on Line-start Permanent Magnet S...

引用

2023 IEEE International Conference on Applied Superconductivity and Electromagnetic Devices, ASEMD 2023

作者： Wang, Yanbo Shen, Tao Xu, Yongming Xu, Ziyi Harbin University of Science and Technology School of Measurement and Communication Engineering Harbin China Changzhou Institute of Technology School of Electrical and Information Engineering Changzhou China

ISBN: (纸本)9798350301571

This paper proposes a torque calculation method based on analytical equations. This method is used to investigate the torque ripple of line-start permanent magnet synchronous motors (LSPMSM). The comparison with the results of the finite element method (FEM) illustrates that the method is able to explore the interaction between magnetic potential and permeance. By calculating torque in two cases, without rotor bar and with rotor bar, the effect of rotor slotting and rotor bar on torque ripple is discussed. The calculation results indicate that rotor slotting is the main source of torque ripple in LSPMSM. © 2023 IEEE.

关键词： Line-start permanent magnet synchronous motors (LSPMSM) rotor slotting torque tipple

来源：评论

学校读者我要写书评

暂无评论

Exploiting the Direct Link in IRS Assisted NOMA Networks with Hardware Impairments

引用

Computer Modeling in engineering & Sciences 2023年第7期136卷 767-785页

作者： Ziwei Liu Xinwei Yue Shuo Chen Xuliang Liu Yafei Wang Wanwei Tang Key Laboratory of Modern Measurement&Control Technology Ministry of EducationSchool of Information and Communication EngineeringBeijing Information Science and Technology UniversityBeijing100101China The Party Committee Office Yellow River Conservancy Technical InstituteKaifeng475001China Tangshan University Tangshan063000China

Hardware impairments(HI)are always present in low-cost wireless *** paper investigates the outage behaviors of intelligent reflecting surface(IRS)assisted non-orthogonal multiple access(NOMA)networks by taking into account the impact of ***,we derive the approximate and asymptotic expressions of the outage probability for the IRS-NOMA-HI *** on the asymptotic results,the diversity orders under perfect self-interference cancellation and imperfect self-interference cancellation scenarios are obtained to evaluate the performance of the considered *** addition,the system throughput of IRS-NOMA-HI is discussed in delay-limited *** obtained results are provided to verify the accuracy of the theoretical analyses and reveal that:1)The outage performance and system throughput for IRS-NOMA-HI outperforms that of the IRS-assisted orthogonal multiple access-HI(IRS-OMA-HI)networks;2)The number of IRS elements,the pass loss factors,the Rician factors,and the value of HI are pivotal to enhancing the performance of IRS-NOMAHI networks;and 3)It is recommended that effective methods of reducing HI should be used to ensure system performance,in addition to self-interference cancellation techniques.

关键词： Hardware impairments imperfect SIC intelligent reflecting surface non-orthogonal multiple access outage probability

来源：评论

学校读者我要写书评

暂无评论

Research on hyperspectral image classification based on improved deep cross-domain few-shot learning 23

Research on hyperspectral image classification based on impr...

引用

Proceedings of the 2023 5th International Conference on Video, Signal and Image Processing

作者： Xinhua Luo Qingyan Wang Xintao Liang Zhen Zhang School of Measurement-Control and Communication Engineering Harbin University of Science and Technology China

ISBN: (纸本)9798400709272

In this study, an enhanced method and system for hyperspectral image classification are presented, based on deep cross-scene few-shot learning. This pertains to the domain of remote sensing image processing technology and addresses the prevalent issues of inadequate classification performance in existing techniques for hyperspectral image categorization. The core aspects of this invention encompass the following: utilization of two mapping layers to standardize the input dimensions between the source and target domains; the deployment of an embedded feature extractor to incorporate the image cubes from both the source and target domains into a space-spectral embedding environment simultaneously, ensuring that like samples are closely aligned and dissimilar ones are distanced. Through gauging the distances between each class of unlabeled and labeled samples in this space-spectral embedding zone, learning with a few number of examples in both the source and target domains is achieved. Furthermore, a conditional domain discriminator is employed to mitigate domain shifts between domains, thus solidifying the domain stability of the extracted spatial-spectral embedding features. This innovative approach allows for high-precision hyperspectral data categorization, even when only a few examples are available.

关键词： Hyperspectral image cross-scene few-shot learning

来源：评论

学校读者我要写书评

暂无评论

Study on Dynamic Behavior of SF6/N2 Discharge Under Narrow-Gap Uniform Electric Field 7

Study on Dynamic Behavior of SF6/N2 Discharge Under Narrow-G...

引用

7th International Conference on Electric Power Equipment - Switching technology, ICEPE-ST 2024

作者： Wang, Yingli Cao, Dongyue Zhang, Mingquan Lu, Qiming Zheng, Dianchun College of Measurement and Communication Harbin University of Science and Technology Harbin China School of Electrical and Electronic Engineering Harbin University of Science and Technology Harbin China

ISBN: (纸本)9798350388947

In this paper, a hydrodynamic model of narrow-gap SF6/N2 gas discharge with flat electrode configuration was established, and the microscopic behavior characteristics of sulfur hexafluoride gas discharge under uniform electric field conditions were strictly studied, and the whole process of discharge triggering, generation of charged particles, particle adhesion, particle recombination, and evolution of discharge trajectory until gap breakdown was revealed. What's more, different discharge triggering conditions were further explored, and the influence mechanism of the initial seed electron density on the model discharge process was investigated in detail. The research results showed that the initial seed electron group density peak N0 was tightly correlated to the spatial charged particles, the gap electric field distribution, and the evolution of the discharge trajectory. Although the initial seed electron group position Z0 has a certain effect on the microscopic behavior of discharge, it is mainly reflected in the spatiotemporal scale of the streamer development process. © 2024 IEEE.

关键词： Electrons

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：