检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Papa, Lorenzo Russo, Paolo Amerini, Irene Zhou, Luping The Department of Computer Control and Management Engineering Sapienza University of Rome 00185 Italy The School of Electrical and Information Engineering Faculty of Engineering The University of Sydney NSW2006 Australia

Vision Transformer (ViT) architectures are becoming increasingly popular and widely employed to tackle computer vision applications. Their main feature is the capacity to extract global information through the self-attention mechanism, outperforming earlier convolutional neural networks. However, ViT deployment and performance have grown steadily with their size, number of trainable parameters, and operations. Furthermore, self-attention's computational and memory cost quadratically increases with the image resolution. Generally speaking, it is challenging to employ these architectures in real-world applications due to many hardware and environmental restrictions, such as processing and computational capabilities. Therefore, this survey investigates the most efficient methodologies to ensure sub-optimal estimation performances. More in detail, four efficient categories will be analyzed: compact architecture, pruning, knowledge distillation, and quantization strategies. Moreover, a new metric called Efficient Error Rate has been introduced in order to normalize and compare models' features that affect hardware devices at inference time, such as the number of parameters, bits, FLOPs, and model size. Summarizing, this paper firstly mathematically defines the strategies used to make Vision Transformer efficient, describes and discusses state-of-the-art methodologies, and analyzes their performances over different application scenarios. Toward the end of this paper, we also discuss open challenges and promising research directions. © 2023, CC BY.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Incremental Localization Algorithm Based on Regularized Iteratively Reweighted Least Square

Incremental Localization Algorithm Based on Regularized Iter...

引用

IEEE International Conference on Smart City/SocialCom/SustainCom (SmartCity)

作者： Xiaoyong Yan Aiguo Song Yu Liu Jian He Ronghui Zhu Remote Measurement and Control Key Lab of Jiangsu Province Southeast University Nanjing China School of Computer Engineering Jinling Institute of Technology Nanjing China School of Intelligence Science and Control Engineering Jinling Institute of Technology Nanjing China

ISBN: (纸本)9781509018949

Considering that incremental localization is influenced by the heteroscedasticity problem caused by cumulative errors and the collinearity problem among nodes, this paper has proposed an incremental localization algorithm with consideration to cumulative error and collinearity problem. Using iteratively reweighted method, the algorithm reduces the influences of error accumulation and avoids collinearity problem between nodes with a regularized method. Simulation experiment results show that compared with the previous incremental localization algorithms the proposed algorithm can not only solve the problem of heteroscedasticity, but also obtain a localization solution with high accuracy. In addition, the method also takes into account the influence of collinearity on localization calculation in the process of locating, thus the method is suitable for different monitoring areas and has high adaptability.

关键词： Estimation Software algorithms Algorithm design and analysis Mathematical model Monitoring Wireless sensor networks Distance measurement

来源：评论

学校读者我要写书评

暂无评论

An Efficient Tolerant-Anisotropic Localization for Large-Scale Wireless Sensor Network

An Efficient Tolerant-Anisotropic Localization for Large-Sca...

引用

International Symposium on Parallel Architectures, Algorithms and Programming (PAAP)

作者： Xiaoyong Yan Aiguo Song Yu Liu Xuanyu Gu Remote Measurement and Control Key Lab of Jiangsu Province Southeast University Nanjing China School of Computer Engineering Jinling Institute of Technology Nanjing China School of Intelligence Science and Control Engineering Jinling Institute of Technology Nanjing China

Multi-hop localization is a common method which is suitable for large-scale application. However, it is usually influenced by the network anisotropy, leading to the instability of the localization performance. In order to reduce the influence of the network anisotropy on the localization accuracy, this paper regards the localization as a regression forecasting process by constructing the mapping relationship between hop-counts and Euclidean distances among nodes. The method can effectively avoid the influence of anisotropy on localization, and it has low computation overhead and high localization accuracy without setting complex parameters. The simulation experimental results show that compared with the previous similar algorithms, the proposed algorithm can obtain a faster localization speed and a higher localization accuracy.

关键词： Wireless sensor networks Euclidean distance Training Anisotropic magnetoresistance Support vector machines Handheld computers Network topology

来源：评论

学校读者我要写书评

暂无评论

Block-Decentralized Damping control of Power Systems using Retrofit control Theory

Block-Decentralized Damping Control of Power Systems using R...

引用

European control Conference (ECC)

作者： Takahiro Kawaguchi Takayuki Ishizaki Aranya Chakrabortty Jun-ichi Imura Department of Systems and Control Engineering Graduate School of Engineering Tokyo Institute of Technology Tokyo Japan Electrical and Computer Engineering North Carolina State University Raleigh NC

ISBN: (数字)9783907144022

ISBN: (纸本)9781728188133

This paper develops a new block-decentralized wide-area control design for improving damping performance of power systems based on retrofit control. Retrofit control is a modular control approach for a stable network system whose subsystems are managed by their corresponding subsystem operators. For power systems, these subsystems refer to individual utility areas, and their operators refer to the utility owners. We first show that the state deflections of the grid model caused by a fault cannot be expressed in the form of a local disturbance, thereby making it a bottleneck for satisfactory control performance. To overcome this difficulty, we propose a method for choosing an appropriate subsystem of interest based on the notion of a design structure matrix (DSM). The DSM illustrates the dependency among the different areas of the power system, and helps the utility owners and system operators to choose an appropriate area. The effectiveness of the proposed method is examined through a numerical simulation with the IEEE 68-bus power system.

关键词： Mathematical model Generators Power system stability Power system dynamics Damping control systems

来源：评论

学校读者我要写书评

暂无评论

Fast Fuzzy Clustering Algorithm for Time Series in Industrial Processes

Fast Fuzzy Clustering Algorithm for Time Series in Industria...

引用

International Conference on computer and Communications (ICCC)

作者： He Ji Jinkuan Wang Yinghua Han Qiang Zhao College of Information Science and Engineering Northeastern University Shenyang China School of Computer and Communication Engineering Northeastern University at Qinhuangdao Qinhuangdao China School of Control Engineering Northeastern University at Qinhuangdao Qinhuangdao China

ISBN: (数字)9781728147437

ISBN: (纸本)9781728147444

Process industry data is often continuously measured to capture fluctuations and changes of a production processes, and this sequentially dataset can be looked as time series. We usually don't know the correlation and classification between these data in advance. Clustering is an effective way to solve this problem, and it have been used in many fields in recent years. There are many different fuzzy clustering methods have been proposed, fuzzy c-means clustering (FCM) and fuzzy c-medoids clustering (FCMdd) are most common algorithm among them. But all of these algorithms use Euclidean distance as the similarity measure and align two times series time-by-time. These methods cannot adapt to the complex industrial environment. In this paper, dynamic time warping (DTW) is selected as the distance of the FCMdd method to adapt to complex industrial production environments, and a type of constraint is introduced to speed up it. The novel clustering method we proposed can not only speed up the clustering algorithm, but also improve the accuracy in some case. Three sets of open source sensor and simulated datasets are selected to evaluate this algorithm and achieved satisfactory results.

关键词： Time series analysis Clustering algorithms Euclidean distance Heuristic algorithms Industries Time measurement Production

来源：评论

学校读者我要写书评

暂无评论

Real Time Sound Source Localization Using von-Mises ResNet

Real Time Sound Source Localization Using von-Mises ResNet

引用

2024 IEEE/SICE International Symposium on System Integration, SII 2024

作者： Bozkurtlar, Mert Yen, Benjamin Itoyama, Katsutoshi Nakadai, Kazuhiro School of Engineering Tokyo Institute of Technology Department of Systems and Control Engineering 2-12-1 Ookayama Meguro-ku Tokyo152-8552 Japan Istanbul Technical University Faculty of Management Department of Industrial Engineering Maçka/Beşiktaş/Harbiye Istanbul34367 Turkey Istanbul Technical University Faculty of Computer and Informatics Engineering Department of Computer Engineering Sariyer/Ayazaga Istanbul34485 Turkey The Nakadai Lab Japan Honda Research Institute Japan Co. Ltd 8-1 Honcho Saitama Wako351-0188 Japan

ISBN: (纸本)9798350312072

This paper addresses the task of learning periodic information using deep neural networks to achieve real-time, environment-independent sound source localization. Previous papers showed phase data is the most significant cue in sound source localization tasks and the proposed vM-B DNN was validated to be able to handle such periodic information using on synthesized data. However, they haven't shown its effectiveness and robustness in realistic use cases. This paper introduces a more complex model based on residual networks and adapts vM-B activation function for convolutional layers for use cases that require real-time predictions in dynamically changing environments. © 2024 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Information sets partition based on entropy using improved particle swarm optimization algorithm

引用

Journal of Computational Information Systems 2010年第7期6卷 2253-2260页

作者： Su, Rijian Kong, Li Liu, Wenzhong Cheng, Jingjing Song, Shengli Guo, Gongbin Dept. of Control Science and Engineering Huazhong University of Science and Technology Wuhan 430074 China School of Computer and Communication Engineering Zhengzhou University of Light Industry Zhengzhou 450002 China

An improved particle swarm optimization (IPSO) with oscillating inertia weight factor and self-adaptation mutation factor is proposed in this paper. The IPSO algorithm is used to achieve the partition of information sets. In the process of partition, the concept of Shannon entropy is introduced. The information set is divided into subsets according to measurement information entropy and the constituted objective function, and its essence is the use of different sensors to obtain a greater amount of information. The method is applied to micro resistivity scanning imaging logging data set partition. Copyright © 2010 Binary Information Press.

关键词： Particle swarm optimization (PSO)

来源：评论

学校读者我要写书评

暂无评论

Analysis of three-dimensional reality modeling algorithm in animation design

Analysis of three-dimensional reality modeling algorithm in ...

引用

2nd International Conference on Precision Mechanical Instruments and Measurement Technology, ICPMIMT 2014

作者： Cai, Bai Chuan Mei, Rong Fang Mei, Jian Guo Yang, Bo School of Software East China Normal University Putuo shanghai 200062 China Department of electronic information and control engineering Yibin Vocational and Technical College Yibin Sichuan 644003 China School of computer and communication engineering University Of Chongqing Shapingba Chongqing 400044 China Chongqing Municipal Health Technical School Banan chongqing 401320 China College of Art and Design Shaanxi University of Science and Technology Shanxi Xian710021 China

ISBN: (纸本)9783038351948

During animation design process, when the overlap of animation graphics is overtopping, inaccurate three-dimensional feature points appear in the established model, which resulting in low fidelity of model. For this drawbacks, a three-dimensional reality animation design modeling based on an optimization algorithm of animation modeling fidelity is proposed. Triangle refinement method is utilized to refine feature points distributed disorderly in the three-dimensional animation model, so as to obtain a three-dimensional animation composed of triangles, according to the method of calculating the intersection of intersecting triangles, optimal triangles can be achieved, i.e. the new three-dimensional coordinate points are acquired. Afterwards, two-dimensional coordinate calculation is processed for the new added points to get the exact coordinates of the point in the three-dimensional animation model, eventually obtain a three-dimensional animation model with high degree of fidelity. © (2014) Trans Tech Publications, Switzerland.

关键词： Three dimensional computer graphics

来源：评论

学校读者我要写书评

暂无评论

Undermodelling Detection with Sign-Perturbed Sums

引用

IFAC-PapersOnLine 2017年第1期50卷 2744-2749页

作者： Carè A. Campi M.C. Csáji B.C. Weyer E. Centrum Wiskunde & Informatica (CWI) Science Park 123 Amsterdam Netherlands Institute for Computer Science and Control (SZTAKI) Hungarian Academy of Sciences (MTA) Kende utca 13-17 Budapest H-1111 Hungary Department of Information Engineering University of Brescia Via Branze 38 Brescia 25123 Italy Department of Electrical and Electronic Engineering Melbourne School of Engineering The University of Melbourne Parkville Melbourne Victoria 3010 Australia

Sign-Perturbed Sums (SPS) is a finite sample system identification method that can build exact confidence regions for the unknown parameters of linear systems under mild statistical assumptions. Theoretical studies of the SPS method have assumed so far that the order of the system model is known to the user. In this paper we discuss the implications of this assumption for the applicability of the SPS method, and we propose an extension that, under mild assumptions, i) still delivers guaranteed confidence regions when the model order is correct, and ii) it is guaranteed to detect, in the long run, if the model order is wrong. © 2017

关键词： confidence regions distribution-free results finite sample results least squares parameter estimation system identification

来源：评论

学校读者我要写书评

暂无评论

Mechatronis: The Output Energy of MEMS Explosive Train 4

Mechatronis: The Output Energy of MEMS Explosive Train

引用

2020 4th International Conference on Energy Material, Chemical engineering and Mining engineering, EMCEME 2020

作者： Tan, Ming Zhang, Liang Zeng, Xiangtao Xia, Wen Sichuan Huachuan Industries Co. Ltd Chengdu China School of Electronics Engineering and Computer Science Peking University Beijing China Special Energy Group Xian Qinghua Co. Xi'an China China Xi'an Satellite Control Center Xi'an China

A micron scale of microelectromechanical systems (MEMS) explosive train with applications in both space exploration and electronic control of safety airbag, the MEMS explosive train is designed to shock initiate a linear shaped charge by accelerating a thin metal plate across a small gap. This paper summarizes the experiment results of Φ0.9×0.6mm micro-charge driven different thickness flyer's velocity by Photonic Doppler Velocimetry (PDV), compared to the calculation result of Gurney formulation, the flyer's velocity of micro-charge driven different thickness lower than the calculation. The velocity-time variation curve and the energy variation with distance provide some information available for micro-explosive train design. © 2021 Published under licence by IOP Publishing Ltd.

关键词： MEMS

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：