检索结果-内蒙古大学图书馆

2016 IEEE Far East NDT New Technology & Application Forum

作者： Qi Ouyang Kaixin Qin Xinglan Zhang Zhi Niu Shuaicheng Hou Laboratory of advanced detection technology and Automation equipment College of AutomationChongqing University College of Computer Science and Engineering Chongqing University of technology

It is significant for improving the quality of medium plate to detect the thickness and surface flaws in intermediate cooling process in real time.A novel EMAT&ECT fusion method is proposed to realize three dimensional(3D) imaging for the surface flaws and thickness of medium plate in this *** is used for the medium plate thickness measurement,and ECT for surface defects *** designed a scanning method for 3D imaging of medium plate,which combines EMAT and *** the same time,the 3D imaging principle of EMAT&ECT fusion testing method is *** by simulation and experiments showed that the method is feasible using EMAT&ECT for medium plate quality detection in intermediate cooling process.

关键词： Sensors Fusion 3D Imaging EMAT ECT NDT

来源：评论

学校读者我要写书评

暂无评论

Scale-adaptive low-resolution person re-identification via learning a discriminating surface 25

Scale-adaptive low-resolution person re-identification via l...

引用

25th International Joint Conference on Artificial Intelligence, IJCAI 2016

作者： Zheng, Wang Hu, Ruimin Yu, Yi Jiang, Junjun Chao, Liang Jinqiao, Wang State Key Laboratory of Software Engineering School of Computer Wuhan University China Digital Content and Media Sciences Research Division National Institute of Informatics Japan School of Computer Science China University of Geosciences China National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences China National Engineering Research Center for Multimedia Software Collaborative Innovation Center of Geospatial Technology Hubei Provincial Key Laboratory of Multimedia and Network Communication Engineering Wuhan University China

Person re-identification, as an important task in video surveillance and forensics applications, has been widely studied. But most of previous approaches are based on the key assumption that images for comparison have the same resolution and a uniform scale. Some recent works investigate how to match low resolution query images against high resolution gallery images, but still assume that the low-resolution query images have the same scale. In real scenarios, person images may not only be with low-resolution but also have different scales. Through investigating the distance variation behavior by changing image scales, we observe that scale-distance functions, generated by image pairs under different scales from the same person or different persons, are distinguishable and can be classified as feasible (for a pair of images from the same person) or infeasible (for a pair of images from different persons). The scale-distance functions are further represented by parameter vectors in the scale-distance function space. On this basis, we propose to learn a discriminating surface separating these feasible and infeasible functions in the scale-distance function space, and use it for reidentifying persons. Experimental results on two simulated datasets and one public dataset demonstrate the effectiveness of the proposed framework.

关键词： Vector spaces

来源：评论

学校读者我要写书评

暂无评论

Numerical analysis near singularities in RBF networks

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2018年第1期19卷

作者： Weili Guo Haikun Wei Yew-Soon Ong Jaime Rubio Hervas Junsheng Zhao Hai Wang Kanjian Zhang Key Laboratory of Measurement and Control of CSE Ministry of Education School of Automation Southeast University Nanjing Jiangsu Province P.R. China School of Computer Science and Engineering Nanyang Technological University Singapore School of Mathematics Science Liaocheng University Liaocheng Shandong Province P.R. China Sobey School of Business Saint Mary's University Halifax Nova Scotia Canada

The existence of singularities often affects the learning dynamics in feedforward neural networks. In this paper, based on theoretical analysis results, we numerically analyze the learning dynamics of radial basis function (RBF) networks near singularities to understand to what extent singularities influence the learning dynamics. First, we show the explicit expression of the Fisher information matrix for RBF networks. Second, we demonstrate through numerical simulations that the singularities have a significant impact on the learning dynamics of RBF networks. Our results show that overlap singularities mainly have influence on the low dimensional RBF networks and elimination singularities have a more significant impact to the learning processes than overlap singularities in both low and high dimensional RBF networks, whereas the plateau phenomena are mainly caused by the elimination singularities. The results can also be the foundation to investigate the singular learning dynamics in deep feedforward neural networks.

关键词： RBF networks deep learning learning dynamics numerical analysis singularity

来源：评论

学校读者我要写书评

暂无评论

Segmentation of ischemic stroke lesions in multi-spectral MR images using weighting suppressed FCM and three phase level set 1st

Segmentation of ischemic stroke lesions in multi-spectral MR...

引用

1st International Workshop on Brainlesion, Brainles 2015 Held in Conjunction with International Conference on Medical Image Computing for computer-Assisted Intervention, MICCAI 2015

作者： Feng, Chaolu Zhao, Dazhe Huang, Min School of Computer Science and Engineering Northeastern University ShenyangLiaoning110819 China Key Laboratory of Medical Image Computing of Ministry of Education Northeastern University ShenyangLiaoning110819 China State Key Laboratory of Synthetical Automation for Process Industries Northeastern University ShenyangLiaoning110819 China

ISBN: (纸本)9783319308579

Accurate segmentation of ischemic lesions is still a challenging task. In this paper, we propose a framework to extract ischemic lesions from multi-spectral MR images. In the proposed framework, MR images of each modality are first segmented into brain tissues and ischemic lesions by weighting suppressed fuzzy c-means. Preliminary lesion segmentation results are then fused among all the imaging modalities by majority voting. The fused segmentation results are finally refined by a three phase level set method. The level set formulation is defined on multi-spectral images with the capability of dealing with intensity inhomogeneities. The proposed framework has been applied to the MICCAI 2015 ISLES challenge. According to the ranking rules of the challenge, the proposed framework took the second place and the fourth place in sub-acute lesion segmentation and acute stroke estimation, respectively. © Springer International Publishing Switzerland 2016.

关键词： Numerical methods

来源：评论

学校读者我要写书评

暂无评论

Cross-modal subspace learning for fine-grained sketch-based image retrieval

arXiv

引用

arXiv 2017年

作者： Xu, Peng Yin, Qiyue Huang, Yongye Song, Yi-Zhe Ma, Zhanyu Wang, Liang Xiang, Tao Bastiaan Kleijn, W. Guo, Jun Pattern Recognition and Intelligent System Laboratory Beijing University of Posts and Telecommunications Beijing China National Lab of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing China SketchX Lab School of Electronic Engineering and Computer Science Queen Mary University of London London United Kingdom Communications and Signal Processing Group Victoria University of Wellington New Zealand

Sketch-based image retrieval (SBIR) is challenging due to the inherent domain-gap between sketch and photo. Compared with pixel-perfect depictions of photos, sketches are iconic renderings of the real world with highly abstract. Therefore, matching sketch and photo directly using low-level visual clues are unsufficient, since a common low-level subspace that traverses semantically across the two modalities is non-trivial to establish. Most existing SBIR studies do not directly tackle this cross-modal problem. This naturally motivates us to explore the effectiveness of cross-modal retrieval methods in SBIR, which have been applied in the image-text matching successfully. In this paper, we introduce and compare a series of state-of-the-art cross-modal subspace learning methods and benchmark them on two recently released fine-grained SBIR datasets. Through thorough examination of the experimental results, we have demonstrated that the subspace learning can effectively model the sketch-photo domain-gap. In addition we draw a few key insights to drive future research. Copyright © 2017, The Authors. All rights reserved.

关键词： Abstracting

来源：评论

学校读者我要写书评

暂无评论

Simulation and experiment study of non-destructive testing using EMA/EC dual probe

Simulation and experiment study of non-destructive testing u...

引用

2016 IEEE Far East NDT New Technology and Application Forum, FENDT 2016

作者： Zhi, Niu Qi, Ouyang Zhixi, Shen Xinlan, Zhang Tianfeng, Zhang Key Laboratory of Dependable Service Computing in Cyber Physical Society of Ministry of Education Chongqing University Chongqing400044 China College of Automation Chongqing University Chongqing400044 China College of Computer Science and Engineering Chongqing University of Technology Chongqing400054 China

ISBN: (纸本)9781509026630

It is of some advantages for non-destructive testing surface and internal flaws using EC/EMA dual probe. The fusion detection principle of EMA/EC is emulated by finite element method. Aimed to EC, the current eddy field distribution focused on skin and coil impedance variation induced by eddy field is studied by theory and experiment method, and aimed to EMAT, the propagation characteristics of the ultrasonic wave excited by EMAT is analyzed by Multiphysics Comsol simulation software. The simulation and experiment results showed that it is effective methods for non-destructive testing surface and internal flaw of the tested object using EC/EMA fusion detection method. © 2016 IEEE.

关键词： Nondestructive examination

来源：评论

学校读者我要写书评

暂无评论

Adaptive Neural Control of Robot Manipulator with Prescribed Performance

Adaptive Neural Control of Robot Manipulator with Prescribed...

引用

第35届中国控制会议

作者： Min Wang Anle Yang School of Automation Science and Engineering Guangzhou Key Laboratory of Brain Computer Interaction and ApplicationsSouth China University of Technology

This paper focuses on adaptive neural control of robot manipulator with unknown system dynamics under the limitation of prescribed performance. A performance function is introduced to express the prescribed constraints of tracking ***, a performance transformation method is proposed to solve the problem of the prescribed performance. The unknown dynamics of robot are approximated by radial basis function(RBF) neural networks(NN). An adaptive neural control law is proposed based on Lyapunov stability. The proposed control method guarantees that all the signals in closed-loop system are uniformly ultimately bounded(UUB), simultaneously the tracking errors converge to small preassigned regions of zero with satisfying prescribed performance. Simulation studies are shown to demonstrate the effectiveness of the proposed control method.

关键词： Adaptive Neural Control Prescribed Performance Performance Transformation Robot Manipulator

来源：评论

学校读者我要写书评

暂无评论

Learning from Adaptive Neural Control of Manipulator Using an ISS-Modular Approach

Learning from Adaptive Neural Control of Manipulator Using a...

引用

第28届中国控制与决策会议

作者： Anle Yang Min Wang School of Automation Science and Engineering Guangzhou Key Laboratory of Brain Computer Interaction and ApplicationsSouth China University of Technology

ISBN: (纸本)9781467397155

This paper studies learning from adaptive neural control of manipulator using an input-to-state stability(ISS) modular *** order to remove the limitation of the derivative of the affine term in robot manipulator,an ISS modular method is introduced into this ***,the proposed adaptive neural control law,based on the ISS modular method and backstepping technology,guarantees the boundedness of all signals in closed-loop system and exponential convergence of tracking ***,by using a state transformation strategy and coupling with the established partial persistent excitation(PE) condition,neural network(NN) weight values converge exponentially to their optimal *** convergent weights which are stored as the constant experienced knowledge make the learning control for a similar task *** results are shown to demonstrate the effectiveness of the proposed control method.

关键词： ISS adaptive neural control robot manipulator PE deterministic learning

来源：评论

学校读者我要写书评

暂无评论

Research on Dust Evolution Simulation System Based on Dynamic Data Driven

Research on Dust Evolution Simulation System Based on Dynami...

引用

2016 International Conference on Sustainable Energy, Environment and Information engineering (SEEIE 2016)

作者： Xiao-mei HU He-wei QU Jun-jun HU Yuan TIAN Shanghai Key Laboratory of Intelligent Manufacturing and Robotics School of Mechatronic Engineering and AutomationShanghai University Electrical Engineering and Computer Science College of EngineeringUniversity of Michigan

The dust evolution simulation provides an important way to analyze the impact of dust on the environment. Dynamic data driven application system framework is applied and dust evolution simulation model based on DDDAS framework is proposed. Compared with the simulation model of dust evolution without DDDAS, the functions of the inputs of real-time data and modification of simulation parameters ensure the improvement of the accuracy of dust evolution simulation. The experimental results show that the dust evolution simulation with DDDAS framework is more precise than the simulation without DDDAS, which proves the effectiveness of the dust evolution simulation system based on dynamic data driven.

关键词： Dynamic data driven Dust evolution simulation Kinetic monte carlo Dust accumulation

来源：评论

学校读者我要写书评

暂无评论

Shape reconstruction within an optical diffusion model

Shape reconstruction within an optical diffusion model

引用

International Workshop on Advanced Motion Control (AMC)

作者： Yangjie Wei Yuqing He State Key Laboratory of Synthetical Automation for Process Industries Shenyang China College of Computer Science and Engineering Northeastern University CAS Shenyang Institute of Automation Shenyang China

Shape reconstruction with an optical diffuser has some significant advantages such as high-precision depth estimation with a small lens, depth estimation for distant objects, and less sensitivity to lens aberrations. However, no optical diffusion model has been proposed to relate depth information to the basic principle of intensity distribution during optical diffusion. In this paper, heat diffusion equation in physics is introduced into optical diffusion, and a mathematic model of intensity distribution during optical diffusion is constructed. A high-precision shape reconstruction method with optical diffusion is proposed based on heat diffusion equation. First, an optical diffusion model is introduced to explain the basic principle of diffusion imaging process. Second, a novel shape reconstruction method based on global heat diffusion is proposed within our optical diffusion model. Finally, simulation with synthetic images and experiment with five playing cards are conducted, and the result proves the effectiveness and feasibility of the proposed method.

关键词： Optical imaging Optical sensors Mathematical model Optical scattering Shape Optical reflection Optical refraction

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：