检索结果-南通市图书馆

Visual attention and clustering-based automatic selection of landmarks using single camera

Journal of Central South University 2014年第9期21卷 3525-3533页

作者： CHUHO yi yONGMIN shin JUNGWON Cho Future IT R&D Lab LG Electronics Home Appliance R&D Lab LG ElectronicsSeoul 153721 Korea Department of Computer Education Jeju National University Department of Mathematics and Computer Science Salisbury University

An improved method with better selection capability using a single camera was presented in comparison with previous method. To improve performance, two methods were applied to landmark selection in an unfamiliar indoor environment. First, a modified visual attention method was proposed to automatically select a candidate region as a more useful landmark. In visual attention, candidate landmark regions were selected with different characteristics of ambient color and intensity in the image. Then, the more useful landmarks were selected by combining the candidate regions using clustering. As generally implemented, automatic landmark selection by vision-based simultaneous localization and mapping(SLAM) results in many useless landmarks, because the features of images are distinguished from the surrounding environment but detected repeatedly. These useless landmarks create a serious problem for the SLAM system because they complicate data association. To address this, a method was proposed in which the robot initially collected landmarks through automatic detection while traversing the entire area where the robot performed SLAM, and then, the robot selected only those landmarks that exhibited high rarity through clustering, which enhanced the system performance. Experimental results show that this method of automatic landmark selection results in selection of a high-rarity landmark. The average error of the performance of SLAM decreases 52% compared with conventional methods and the accuracy of data associations increases.

关键词： simultaneous localization and mapping automatic landmark selection visual attention clustering

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Tensor based non-photo realistic rendering for korean architectural images

引用

Journal of Central South University 2014年第10期21卷 3805-3810页

作者： Do kyung shin Eun young Ahn Deprtment of Computer Science and Engineering Hanyang University Deprtment of Communication Infromation & Computer Engineering Hanbat National University

Existing pen and ink sketch technologies can be applied to general images, but they could not produce optimal output for images of traditional architecture, because most images consist of exquisite straight lined patterns in traditional architecture, such as root tiles and window bars. The lines of roofs and eaves need to be described delicately to express pen and ink sketch most effectively. Therefore, by proposing a method to create white noise for light and shade of input images, to extract input vector directions from the white noise, and to determine the direction and length of stroke, a new expression technique is proposed for pen and ink sketch that could best reflect the characteristics of traditional architecture.

关键词： non-photo realistic rendering pen and ink sketch tensor

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Feature-aware regularization for sparse online learning

引用

Science China(Information Sciences) 2014年第5期57卷 65-85页

作者： OIWA Hidekazu MATSUSHIMA shin NAkAGAWA Hiroshi Graduate School of Information Science and Technology The University of Tokyo Information Technology Center The University of Tokyo

Learning a compact predictive model in an online setting has recently gained a great deal of *** combination of online learning with sparsity-inducing regularization enables faster learning with a smaller memory space than the previous learning *** optimization methods and learning algorithms have been developed on the basis of online learning with L1-regularization.L1-regularization tends to truncate some types of parameters,such as those that rarely occur or have a small range of values,unless they are emphasized in ***,the inclusion of a pre-processing step would make it very difficult to preserve the advantages of online *** propose a new regularization framework for sparse online *** focus on regularization terms,and we enhance the state-of-the-art regularization approach by integrating information on all previous subgradients of the loss function into a regularization *** resulting algorithms enable online learning to adjust the intensity of each feature’s truncations without pre-processing and eventually eliminate the bias of *** show theoretical properties of our framework,the computational complexity and upper bound of *** demonstrated that our algorithms outperformed previous methods in many classification tasks.

关键词： online learning supervised learning sparsity-inducing regularization feature selection sentiment analysis

来源：

同方期刊数据库评论

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Video learning based image classification method for object recognition

引用

Journal of Central South University 2013年第9期20卷 2399-2406页

作者： LEE Hong-ro shin yong-ju Department of NTIS Center Korea Institute of Science & Technology Information

Automatic image classification is the first step toward semantic understanding of an object in the computer vision *** key challenge of problem for accurate object recognition is the ability to extract the robust features from various viewpoint images and rapidly calculate similarity between features in the image database or video *** order to solve these problems,an effective and rapid image classification method was presented for the object recognition based on the video learning *** optical-flow and RANSAC algorithm were used to acquire scene images from each video *** the selection of scene images,the local maximum points on comer of object around local area were found using the Harris comer detection algorithm and the several attributes from local block around each feature point were calculated by using scale invariant feature transform （SIFT） for extracting local ***,the extracted local descriptor was learned to the three-dimensional pyramid match *** results show that our method can extract features in various multi-viewpoint images from query video and calculate a similarity between a query image and images in the database.

关键词： image classification multi-viewpoint image feature extraction video learning

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Existence and Uniqueness of Solutions for two Classes of Functional Equations Arising in Dynamic Programming

引用

Acta Mathematicae Applicatae Sinica 2007年第2期23卷 195-208页

作者： Ze-qing Liu shin Min kang Department of Mathematics Liaoning Normal University P.O. Box 200 Dalian 116029 China Department of Mathematics and The Research Institute of Natural Science Gyeongsang National UniversityChinju 660-701 Korea

In this paper we establish the existence, uniqueness and iterative approximation of solutions for two classes of functional equations arising in dynamic programming of multistage decision processes. The results presented here extend, and unify the corresponding results due to Bellman, Bhakta and Choudhury, Bhakta and Mitra, Liu and others.

关键词： Functional equation dynamic programming fixed point successive approximation nonexpansive mapping

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Learning excited states from ground states by using an artificial neural network

引用

npj Computational Materials 2020年第1期6卷 1056-1061页

作者： shin kiyohara Masashi Tsubaki Teruyasu Mizoguchi Institute of Industrial Science University of Tokyo153-8505 TokyoJapan National Institute of Advanced Industrial Science and Technology(AIST) 135-0064 TokyoJapan

Excited states are different quantum states from their ground states,and spectroscopy methods that can assess excited states are widely used in materials *** the spectra reflecting excited states is thus of great importance for materials ***,understanding such spectra remains difficult because excited states have usually different atomic or electronic configurations from their corresponding ground *** excited states could be predicted from ground states,the knowledge of the excited states would be ***,we used an artificial neural network to predict the excited states of the core-electron absorption spectra from their ground ***,our model correctly learned and predicted the excited states from their ground states,providing several thousand times computational ***,it showed excellent transferability to other ***,we found two physical insights about excited states:core-hole effects of amorphous silicon oxides are stronger than those of crystalline silicon oxides,and the excited-ground states relationships of some metal oxides are similar to those of the silicon oxides,which could not be obtained by conventional spectral simulation nor found until using machine leaning.

关键词： ground artificial oxides

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Least-squares finite-element method for shallow-water equations with source terms

引用

Acta Mechanica Sinica 2009年第5期25卷 597-610页

作者： shin-Jye Liang Tai-Wen Hsu Department of Marine Environmental lnformatics National Taiwan Ocean University Department of Hydraulic and Ocean Engineering National Cheng Kung University

Numerical solution of shallow-water equations （SWE） has been a challenging task because of its nonlinear hyperbolic nature, admitting discontinuous solution, and the need to satisfy the C-property. The presence of source terms in momentum equations, such as the bottom slope and friction of bed, compounds the difficulties further. In this paper, a least-squares finite-element method for the space discretization and θ-method for the time integration is developed for the 2D non-conservative SWE including the source terms. Advantages of the method include： the source terms can be approximated easily with interpolation functions, no upwind scheme is needed, as well as the resulting system equations is symmetric and positive-definite, therefore, can be solved efficiently with the conjugate gradient method. The method is applied to steady and unsteady flows, subcritical and transcritical flow over a bump, 1D and 2D circular dam-break, wave past a circular cylinder, as well as wave past a hump. Computed results show good C-property, conservation property and compare well with exact solutions and other numerical results for flows with weak and mild gradient changes, but lead to inaccurate predictions for flows with strong gradient changes and discontinuities.

关键词： Shallow-water equations Source terms Least-squares finite-element method Dam-break C-property

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Implementation of Turbulent Mixing over a Stratocumulus-Topped Boundary Layer and Its Impact in a GCM

引用

Advances in Atmospheric Sciences 2009年第5期26卷 995-1004页

作者： Sun-Hee shin kyung-Ja HA Division of Earth Environmental System Pusan National University

The effect of a vertical diffusion scheme over a stratocumulus topped boundary layer （STBL） was investigated using the yONU AGCM （yonsei University Atmospheric General Circulation Model）. To consider the impact of clouds on the turbulence production, the turbulence mixing term, driven by radiative cooling at the cloud top, is implemented as an extended non-local diffusion scheme. In the model with this new scheme, the STBL parameterization significantly influences the lower atmosphere over the tropical and subtropical regions. Consideration of the turbulent mixing within the cloud layer leads to continuous stratocumulus formation. The cloud-top radiative cooling tends to favor more rapid entrainment and produces top-down turbulent mixing. This cooling develops a mixed layer without initiation of deep convection by surface fluxes. Variations in thermodynamical and dynamical features are produced by planetary boundary layer （PBL） cloud development. The simulated stratocumulus induces more mixing of heat and moisture due to the cloud forcing. Over STBL regions, the lower boundary layer becomes warmer and drier. It also weakens vertical motion and zonal trade winds in the eastern Pacific, which indicates that stratocumulus cloud cover plays a role in weakening the Walker circulation; that is, cloud cover damps the tropical circulation.

关键词： stratocumulus topped boundary layer non-local diffusion scheme tropical circulation general circulation model

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

APPROXIMATION OF COMMON SOLUTIONS OF VARIATIONAL INEQUALITIES VIA STRICT PSEUDOCONTRACTIONS

引用

Acta Mathematica Scientia 2012年第4期32卷 1607-1618页

作者： Sun young Cho shin Min kang Department of Mathematics and RINS Gyeongsang National UniversityJinju 660-701 Korea

In this paper, a convex feasibility problem is considered. We construct an iterative method to approximate a common element of the solution set of classical variational inequalities and of the fixed point set of a strict pseudocontraction. Strong convergence theorems for the common element are established in the framework of Hilbert spaces.

关键词： fixed point inverse-strongly monotone mapping nonexpansive mapping variational inequality

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Use of Local Region Maps on Convolutional LSTM for Single-Image HDR Reconstruction

引用

Computers, Materials & Continua 2022年第6期71卷 4555-4572页

作者： Seungwook Oh GyeongIk shin Hyunki Hong College of Software Chung-Ang UniversityHeukseok-ro 84Dongjak-kuSeoul06973Korea

Low dynamic range(LDR)images captured by consumer cameras have a limited luminance *** the conventional method for generating high dynamic range(HDR)images involves merging multiple-exposure LDR images of the same scene(assuming a stationary scene),we introduce a learning-based model for single-image HDR *** input LDR image is sequentially segmented into the local region maps based on the cumulative histogram of the input brightness *** the local region maps,SParam-Net estimates the parameters of an inverse tone mapping function to generate a pseudo-HDR *** process the segmented region maps as the input sequences on long short-term ***,a fast super-resolution convolutional neural network is used for HDR image *** proposed method was trained and tested on datasets including HDR-Real,LDR-HDR-pair,and *** experimental results revealed that HDR images can be generated more reliably than using contemporary end-to-end approaches.

关键词： Low dynamic range high dynamic range deep learning convolutional long short-term memory inverse tone mapping function

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

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：