检索结果-内蒙古大学图书馆

Exploiting Inter-Sample Affinity for Knowability-Aware Universal Domain Adaptation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wang, Yifan Zhang, Lin Song, Ran Li, Hongliang Rosin, Paul L. Zhang, Wei School of Control Science and Engineering Shandong University Jinan China School of Information and Communication Engineering University of Electronic Science and Technology of China Chengdu China School of Computer Science and Informatics Cardiff University Cardiff United Kingdom

Universal domain adaptation (UniDA) aims to transfer the knowledge of common classes from the source domain to the target domain without any prior knowledge on the label set, which requires distinguishing in the target domain the unknown samples from the known ones. Recent methods usually focused on categorizing a target sample into one of the source classes rather than distinguishing known and unknown samples, which ignores the inter-sample affinity between known and unknown samples, and may lead to suboptimal performance. Aiming at this issue, we propose a novel UniDA framework where such inter-sample affinity is exploited. Specifically, we introduce a knowability-based labeling scheme which can be divided into two steps: 1) Knowability-guided detection of known and unknown samples based on the intrinsic structure of the neighborhoods of samples, where we leverage the first singular vectors of the affinity matrix to obtain the knowability of every target sample. 2) Label refinement based on neighborhood consistency to relabel the target samples, where we refine the labels of each target sample based on its neighborhood consistency of predictions. Then, auxiliary losses based on the two steps are used to reduce the inter-sample affinity between the unknown and the known target samples. Finally, experiments on four public datasets demonstrate that our method significantly outperforms existing state-of-the-art methods. Copyright © 2022, The Authors. All rights reserved.

关键词： Machine learning

A Static Feature Point Extraction Algorithm for Visual-Inertial SLAM

学校读者我要写书评

暂无评论

A Static Feature Point Extraction Algorithm for Visual-Inert...

Chinese Automation Congress (CAC)

作者： HanXuan Zhang Ju Huo Wei Sun Muyao Xue JianBao Zhou School of Electrical Engineering and Automation Harbin Institute of Technology Harbin China Shanghai Space Propulsion Technology Research Institute Shanghai China Harbin Nuoxin University of Technology Measurement and Control Technology Co. Ltd Harbin China

Aiming at the fact that most simultaneous localization and mapping (SLAM) methods are fragile in dynamic environment, a static feature point extraction algorithm for visual-inertial SLAM is proposed. Firstly, the distance from feature matching to transformation matrix is calculated by cleverly combining inertial measurement unit (IMU) data and visual information. Secondly, a static probability model for distinguishing feature point categories is established based on the binomial logistic regression model. These categorical feature points are used in subsequent algorithms to select more stable data associations. Finally, the algorithm is integrated into the ORBSLAM-3 system, and extensive experiments are performed on the OpenLORIS-Scene public dataset and real world to verify the performance of the integrated system. The results show our algorithm not only maintains the high-precision localization performance of the original system in the static environment, but also improves the localization accuracy of the original system in the dynamic environment.

关键词： Location awareness Visualization Simultaneous localization and mapping measurement units Automation Heuristic algorithms Inertial navigation

Additive manufacturing of silicon carbide microwave-absorbing metamaterials

学校读者我要写书评

暂无评论

Additive Manufacturing Frontiers 2025年第1期4卷

作者： Zhao, Hanqing Liao, Qingwei Li, Yinghao Chu, Xiangcheng Yuan, Songmei Qin, Lei Key Laboratory of Sensors Beijing Information Science and Technology University Beijing 100192 China Key Laboratory of Modern Measurement & Control Technology Ministry of Education Beijing Information Science & Technology University Beijing 100192 China State Key Laboratory of New Ceramics and Fine Processing School of Material Science and Engineering Tsinghua University Beijing 100084 China School of Mechanical Engineering and Automation Beihang University Beijing 100191 China

SiC is a wave-absorbing material with good dielectric properties, high-temperature resistance, and corrosion resistance, which has great potential for development in the field of high-temperature wave-absorbing. However, SiC is limited by its low impedance-matching performance and single wave-absorbing mechanism. Therefore, compatible metamaterial technologies are required to enhance its wave-absorbing performance further. The electromagnetic wave (EMW) absorbing metamaterials can realize perfect absorption of EMWs in specific frequency bands and precise regulation of EMW phase, propagation mode, and absorption frequency bands through structural changes. However, the traditional molding methods for manufacturing complex geometric shapes require expensive molds, involve process complexity, and have poor molding accuracy and other limitations. Therefore, additive manufacturing (AM) technology, through material layered stacking to achieve the processing of materials, is a comprehensive multidisciplinary advanced manufacturing technology and has become the core technology for manufacturing metamaterials. This review introduces the principles and applications of different AM technologies for SiC and related materials, discusses the current status and development trends of various AM technologies for fabricating silicon–carbon-based wave-absorbing metamaterials, summarizes the limitations and technological shortcomings of existing AM technologies for fabricating silicon–carbon-based wave-absorbing metamaterials, and provides an outlook for the future development of related AM technologies. © 2024 The Author(s)

关键词： Additive manufacturing Electromagnetic absorption Metamaterials SiC

Wind Power Forecasting Methods Based on Deep Learning:A Survey

学校读者我要写书评

暂无评论

Computer Modeling in engineering & Sciences 2020年第1期122卷 273-301页

作者： Xing Deng Haijian Shao Chunlong Hu Dengbiao Jiang Yingtao Jiang School of Computer Science Jiangsu University of Science and TechnologyZhenjiangChina School of Automation Key Laboratory of Measurement and Control for CSEMinistry of EducationSoutheast UniversityNanjingChina Department of Electrical and Computer Engineering University of NevadaLas VegasUSA

Accurate wind power forecasting in wind farm can effectively reduce the enormous impact on grid operation safety when high permeability intermittent power supply is connected to the power *** to provide reference strategies for relevant researchers as well as practical applications,this paper attempts to provide the literature investigation and methods analysis of deep learning,enforcement learning and transfer learning in wind speed and wind power forecasting ***,wind speed and wind power forecasting around a wind farm requires the calculation of the next moment of the definite state,which is usually achieved based on the state of the atmosphere that encompasses nearby atmospheric pressure,temperature,roughness,and *** an effective method of high-dimensional feature extraction,deep neural network can theoretically deal with arbitrary nonlinear transformation through proper structural design,such as adding noise to outputs,evolutionary learning used to optimize hidden layer weights,optimize the objective function so as to save information that can improve the output accuracy while filter out the irrelevant or less affected information for *** establishment of high-precision wind speed and wind power forecasting models is always a challenge due to the randomness,instantaneity and seasonal characteristics.

关键词： Deep learning reinforcement learning transfer learning wind power forecasting

Joint beamforming and compressed sensing for uplink grant-free access

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Xia, Guoqing Xiao, Pei Li, Bohan Zhang, Yue Zhou, Huiyu School of Engineering University of Leicester LeicesterLE1 7RH United Kingdom University of Surrey GuildfordGU2 7XH United Kingdom School of Computing and Mathematical Sciences University of Leicester LeicesterLE1 7RH United Kingdom Institute of Communication Measurement Technology Chengdu610095 China

Compressed sensing (CS)-based techniques have been widely applied in the grant-free non-orthogonal multiple access (NOMA) to a single-antenna base station (BS). In this paper, we consider the multi-antenna reception at the BS for uplink grant-free access for the massive machine type communication (mMTC) with limited channel resources. To enhance the overloading performance of the BS, we develop a general framework for the synergistic amalgamation of the spatial division multiple access (SDMA) technique with the CS-based grant-free NOMA. We derive a closed-form statistical beamforming and a dynamic beamforming scheme for the inter-cluster interference suppression when applying SDMA. Based on this, we further develop a joint adaptive beamforming and subspace pursuit (J-ABF-SP) algorithm for the multiuser detection and data recovery, with a novel sparsity level decision method without the accurate knowledge of the noise level. To further improve the data recovery performance, we propose an interference cancellation-based J-ABF-SP scheme (J-ABF-SP-IC) by using the initial signal estimates generated from the J-ABF-SP algorithm. Illustrative simulations verify the superior user detection and signal recovery performance of our proposed algorithms in comparison with existing CS-based grant-free NOMA techniques. Copyright © 2025, The Authors. All rights reserved.

关键词： Beamforming

Learning Hierarchical Color Guidance for Depth Map Super-Resolution

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Cong, Runmin Sheng, Ronghui Wu, Hao Guo, Yulan Wei, Yunchao Zuo, Wangmeng Zhao, Yao Kwong, Sam Institute of Information Science Beijing Jiaotong University Beijing100044 China School of Control Science and Engineering Shandong University Jinan250061 China Key Laboratory of Machine Intelligence and System Control Ministry of Education Shandong Jinan250061 China Artificial Intelligence Institute Beijing Normal University Beijing100875 China Changsha410073 China School of Electronics and Communication Engineering Sun Yat-sen University Guangzhou510275 China School of Computer Science and Technology Harbin Institute of Technology Harbin150001 China Lingnan University Hong Kong

Color information is the most commonly used prior knowledge for depth map super-resolution (DSR), which can provide high-frequency boundary guidance for detail restoration. However, its role and functionality in DSR have not been fully developed. In this paper, we rethink the utilization of color information and propose a hierarchical color guidance network to achieve DSR. On the one hand, the low-level detail embedding module is designed to supplement high-frequency color information of depth features in a residual mask manner at the low-level stages. On the other hand, the high-level abstract guidance module is proposed to maintain semantic consistency in the reconstruction process by using a semantic mask that encodes the global guidance information. The color information of these two dimensions plays a role in the front and back ends of the attention-based feature projection (AFP) module in a more comprehensive form. Simultaneously, the AFP module integrates the multi-scale content enhancement block and adaptive attention projection block to make full use of multi-scale information and adaptively project critical restoration information in an attention manner for DSR. Compared with the state-of-the-art methods on four benchmark datasets, our method achieves more competitive performance both qualitatively and quantitatively. The code and results can be found from the link of https://***/HCGNet_TIM2024. © 2024, CC BY-NC-SA.

关键词： Restoration

Coverage Optimization of Camera Network for Continuous Deformable Object

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Chang Chen, Xi Chai, Li The School of Information Science and Engineering Wuhan University of Science and Technology Hubei 430081 China The Engineering Research Center of Metallurgical Automation and Measurement Technology Wuhan University of Science and Technology Hubei 430081 China The College of Control Science and Engineering Zhejiang University Hangzhou310027 China

In this paper, a deformable object is considered for cameras deployment with the aim of visual coverage. The object contour is discretized into sampled points as meshes, and the deformation is represented as continuous trajectories for the sampled points. To reduce the computational complexity, some feature points are carefully selected representing the continuous deformation process, and the visual coverage for the deformable object is transferred to cover the specific feature points. In particular, the vertexes of a rectangle that can contain the entire deformation trajectory of every sampled point on the object contour are chosen as the feature points. An improved wolf pack algorithm is then proposed to solve the optimization problem. Finally, simulation results are given to demonstrate the effectiveness of the proposed deployment method of camera network. © 2022, CC BY-NC-SA.

关键词： Deformation

Enhancement of localization accuracy based on research of conversion coefficient in coherent Φ-OTDR

学校读者我要写书评

暂无评论

TechRxiv

TechRxiv 2023年

作者： Zhong, Zhen Zou, Ningmu Zhang, Xuping You, Chen Li, Meng Department of Measurement-Control Technology Changzhou Institute of Technology Changzhou213032 China Key Laboratory of Intelligent Optical Sensing and Manipulation Ministry of Education Nanjing University Nanjing210093 China Ingram School of Engineering Texas State University San MarcosTX78666 United States

Recently, a distributed fiber optic sensor named phase optical time domain reflectometer or phase-sensitive optical time domain reflectometer (ɸ-OTDR) has been used in the localization and monitoring of earthquake waves, underwater facilities, etc. Precision in the localization of external perturbations on the fiber is crucial for effective problem resolution. The conversion coefficient between the extracted phase signal detected by ɸ-OTDR and the corresponding perturbation signal acting on the fiber significantly impacts localization accuracy. Therefore, the characteristic of parameters relative to the conversion coefficient in ɸ- OTDR should be deeply researched. Then, based on the established mathematical model of coherent ɸ-OTDR, modulus, statistical phase, phase change, and peak difference of coherent ɸ-OTDR with and without the static region is analyzed, respectively. The analysis reveals that when coherent φ-OTDR has perturbations homogeneously distributed along the fiber, there is no static region on the phase change-fiber length plane. At this point, although the waveform like the information of external perturbation still can be found in the retrieved phase change in the direction of pulse sequence, the phase change does not have a strictly linear relationship along the fiber. It is different from what was expected in coherent ɸ-OTDR having the static region, which means that coherent ɸ-OTDR must have the static region if a higher localization accuracy is needed. A localization case demonstrates that the proposed method with the designed static region enhances localization resolution by 210% compared to the traditional coherent φ-OTDR without the static region. These findings underscore the importance of the conversion coefficient and the relevance of the static region in ɸ-OTDR for achieving accurate and effective localization. © 2023, CC BY.

关键词： Fibers

Efficient and Low-Cost Monitoring of Sea Ice Thickness Using Shipborne Vibration Sensors and Machine Learning: A Novel Approach

学校读者我要写书评

暂无评论

Efficient and Low-Cost Monitoring of Sea Ice Thickness Using...

IEEE International Conference on Cyber technology in Automation, control, and Intelligent Systems

作者： Han Zhang Xiyu Quan Le Yang Yuefang Li Leng Han Dongli Deng Weishi Li Xiaoyong Lv Yuliang Zhao Detachment 60 PLA Unit 92493 Huludao China Hebei Key Laboratory of Micro-Nano Precision Optical Sensing and Measurement Technology School of Control Engineering Northeastern University at Qinhuangdao Qinhuangdao China Experimentalist with Northeastern University at Qinhuangdao Qinhuangdao China

Sea ice is a significant factor in ship navigation, and accurately identifying its thickness is crucial for timely decision-making to ensure ship structural integrity and personnel safety. However, current methods for sea ice thickness measurement, such as shipborne imaging devices and synthetic aperture radar altimeters, have limitations such as high costs, difficulties in data acquisition, and poor timeliness. In this article, we present an approach to detect sea ice thickness using a combination of sensors and machine learning techniques. Vibration signals from ship-ice collisions are collected using acceleration vibration sensors, undergo data preprocessing and time-domain alignment, and the true value of sea ice thickness is obtained through field measurements and a sea ice image information identification system. Time-frequency domain features are extracted, and feature quantity optimization is performed, followed by machine learning regression algorithms to predict sea ice thickness. Experimental results show that excellent regression results can be obtained using a single node for thickness prediction, with an R2 value of 0.908 and an RMSE of 1.374 cm. Our findings demonstrate that analyzing vibration signals generated by ship-ice collisions during vessel navigation can effectively identify sea ice thickness. This approach provides a practical and cost-effective alternative for sea ice thickness measurement and identification in the field.

关键词：