检索结果-内蒙古大学图书馆

Chinese Conference on Pattern Recognition

作者： Huang, Xiangsheng Gong, Lujin Wang, Xiaoyan Institute of Automation Chinese Academy of Sciences Beijing 100190 China State Key Laboratory of Scientific and Engineering Computing Institute of Computational Mathematics and Scientific/Engineering Computing AMSS Beijing 100190 China College of Computer Science and Technology Zhejiang University of Technology Hangzhou 310023 China

ISBN: (纸本)9781424472109

Fast facial points fitting plays an important role in applications such as Human-computer Interaction, entertainment, surveillance, and is highly relevant to the techniques of facial expression analysis, face recognition, 3D face model generation, etc. Active Appearance Models (AAMs) are generative models commonly used to fit face. They are sensitive to illumination and expression changes because they use only raw intensity to build observation models. In this paper, a real time facial points fitting approach using mixture observation models is presented. Furthermore, the 3D modes are used to constrain the AAM so that it can only generate model instances that can also be generated with the 3D modes. Finally, we give a derivative process for fast energy minimization using the inverse compositional algorithm. A coarse-to-fine fitting strategy is used for realtime and robust facial points fitting. We apply this algorithm to facial expression cloning of 3D Avatar system. Experimental results demonstrate that fitting the AAM with mixture observation models and 3D constraint outperforms other classical algorithms. ©2010 IEEE.

关键词： Mixtures

来源：评论

学校读者我要写书评

暂无评论

Logical optimization of dataflows for data mining and integration processes

Proceedings - 6th IEEE International Conference on e-Science...

引用

Proceedings - 6th IEEE International Conference on e-science Workshops, e-scienceW 2010 2010年 117-122页

作者： Wöhrer, Alexander Mehofer, Eduard Brezany, Peter University of Vienna Faculty of Computer Science Department of Scientific Computing Vienna Austria

ISBN: (纸本)9780769542959

Modern scientific collaborations require large-scale data mining and integration processes. Their investigations involve multi-disciplinary expertise and large-scale computational experiments on top of large amounts of data that are located in distributed data repositories running various software systems, and managed by different organizations. Higher-level dataflow languages are used on top of parallel dataflow systems to enable faster program development and more maintainable code. Logical and physical optimization should be applied prior to its execution to improve performance. In this paper we present the rationale, theory, design and application of logical optimization of dataflows for data mining and integration processes. A dataflow model is defined and several optimization algorithms, namely dead elements elimination, process re-ordering, parallelization, and data by-passing are developed. The first research prototype of the framework has been implemented in the context of the ADMIRE Data Mining and Integration Process Designer for logical optimization of specifications expressed in the DISPEL language developed in the ADMIRE project. © 2010 IEEE.

关键词： Data integration

来源：评论

学校读者我要写书评

暂无评论

Design issues for high performance simulation

Design issues for high performance simulation

引用

作者： Sloot, P.M.A. Pimentel, A.D. Hertzberger, L.O. Parallel Scientific Computing and Simulation Group Department of Computer Science University of Amsterdam Kruislaan 403 1098 SJ Amsterdam Netherlands Computer Architecture and Parallel Systems Group Department of Computer Science University of Amsterdam Kruislaan 403 1098 SJ Amsterdam Netherlands

One of the key issues in designing new simulation models for parallel execution, or in the migration of existing models to parallel platforms, is the mapping of the application architecture to the parallel system architecture. In this mapping process, we can easily loose track of the inherent locality present in the different architecture layers. In this paper, we present an overview of these issues and examine, by means of several case-studies, the consequences of the design and implementation choices for the various mapping processes. We will show that the potential for high performance simulation comes from a holistic approach, taking into account all aspects from the application to the underlying hardware. © 1998 Elsevier science B. V. All rights reserved.

关键词： Parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

Achievable sum-rate maximizing af relay beamforming scheme in two-way relay channels

Achievable sum-rate maximizing af relay beamforming scheme i...

引用

ICC 2008 - 2008 IEEE International Conference on Communications Workshops

作者： Lee, Namyoon Yang, Hyun Jong Chun, Joohwan Scientific Computing Laboratory School of Electrical Engineering and Computer Science Korea Advanced Institute of Science and Technology 373-1 Guseong-dong Yuseong-gu Daejeon 305-701 Korea Republic of

ISBN: (纸本)9781424420513

This paper deals with design of a amplify-and-forward (AF) relay beamforming matrix in a two-way relay protocol when a relay node has multiple antennas. Two-way relaying is a promising protocol since it provides enhanced spectral efficiency compared to one-way relaying protocols. In this paper we derive the achievable sum-rate upper bound of AF beamforming scheme and propose the achievable sum-rate maximizing relay beamforming scheme when the destination and the relay node have perfect knowledge of the channel state information (CSI) for forward and backward channels. The solution of the sum-rate maximizing scheme is obtained by using the proposed algorithm, general power iterative algorithm, which solves a global optimization problem efficiently. Comparing with one-way relaying protocol, the improved sum-rate performance of the proposed scheme is verified by numerical simulation. ©2008 IEEE.

关键词： Channel state information

来源：评论

学校读者我要写书评

暂无评论

The versioning system balancing data amount and access frequency on distributed storage system

The versioning system balancing data amount and access frequ...

引用

21st International Conference on Data Engineering Workshops 2005

作者： Nakano, Mana Kobayashi, Dai Watanabe, Akitsugu Uehara, Toshihiro Taguchi, Ryo Yokota, Haruo Department of Computer Science Tokyo Institute of Technology Japan Science and Technical Research Laboratories Japan Broadcasting Corporation Japan Global Scientific Information and Computing Center Tokyo Institute of Technology Japan

ISBN: (纸本)0769526578

In this paper, a method of balancing both access frequency and data amount for a distributed parallel storage system under version management is discussed. We assume the version management method keeps the latest version of file and a number of differential information sets to access previous versions. Since the access frequency for an aged version is tend to be lower than that for the latest version, we control the access frequency distribution by the placement of the latest versions, while the data amount distribution is managed by the placement of the differential information whose size is enough small to adjust the subtle difference of data amount. We propose a distributed directory structure and data placement algorithm, and evaluate its effect on the access frequency and data amount distribution. © 2005 IEEE.

关键词： Distributed database systems

来源：评论

学校读者我要写书评

暂无评论

Linear precoder and decoder design for two-way AF MIMO relaying system

Linear precoder and decoder design for two-way AF MIMO relay...

引用

2008 IEEE 67th Vehicular Technology Conference-Spring, VTC

作者： Namyoon, Lee Heesun, Park Joohwan, Chun Scientific Computing Laboratory School of Electrical Engineering and Computer Science Korea Advanced Institute of Science and Technology 373-1 Guseong-dong Yuseong-gu Daejeon 305-701 Korea Republic of

ISBN: (纸本)9781424416455

A relay based wireless network system is basically composed of three nodes, two source nodes which are also destination nodes and the relay node which helps reliable communications between two nodes. The amplify-and-forward (AF) relaying scheme is more convenient to implement than the decode-and-forward (DF) relaying scheme which requires a decoding process at the relay node. In this paper, we design an optimal linear precoder and decoder which minimize the sum of mean squared error (SMSE) when the relay and the destination node have perfect knowledge of the channel state information (CSI) for the forward and the backward channels, respectively. The optimal linear precoder and decoder solutions are obtained by using the proposed iterative algorithm which satisfies the KKT optimality conditions. The average BER performance of the proposed scheme is compared with those of a conventional AF-ZF scheme and AF-MMSE scheme. © 2008 IEEE.

关键词： Channel state information

来源：评论

学校读者我要写书评

暂无评论

Prediction analysis for Parkinson disease using multiple feature selection & classification methods

引用

Multimedia Tools and Applications 2023年第27期82卷 42995-43012页

作者： Hema, M.S. Maheshprabhu, R. Reddy, K. Sudheer Guptha, M. Nageswara Pandimurugan, V. Computer Science and engineering Sri Venkateshwara college of engineering Bengaluru India Mechanical Engineering Aurora Scientific and Technological Institute Hyderabad India Information Technology Anurag University Hyderabad India School of Computing Networking and Communications SRMIST Kattankulathur Chennai India

Diagnosis of Parkinson's disease (PD) is a difficult undertaking that requires a variety of testing and clinical trials. Despite all these testing and trials, there is a considerable risk of misdiagnosis of Parkinson's disease, which causes a delay in making decisions about the best treatment for the patients. computer aided diagnosis can improve the experiences and helps the doctors to predict PD patients. Four sorts of classification algorithms with various types of feature selection approaches are proposed in this paper for successful PD diagnosis. Support Vector Machine-SVM, Nave Bayes-NB, K-Nearest Neighbor-KNN, and Random Forest-RF are the categorization techniques employed. All the dataset's features are not necessary for classification. To choose the relevant characteristics, four feature selection approaches are used. Least Absolute Shrinkage and Selection Operator (LASSO), Backward-forward, rough set, and tree-based feature selection techniques are used. Experimentation is carried out using the Parkinson's Progressive Markers Initiative (PPMI) dataset. Four classification algorithms are constructed, together with and without feature selection procedures, and comparative research is conducted. To measure the performance of the classifiers, various evaluation methodologies are used. Overall Random Forest classifier with four feature selection methods’ gives best results with average accuracy is 96% and rough set feature selection for all three classifiers’ average accuracy is 97%. © 2023, The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Performance characterization of a hierarchical MPI implementation on large-scale distributed-memory platforms

Performance characterization of a hierarchical MPI implement...

引用

38th International Conference on Parallel Processing, ICPP-2009

作者： Alam, Sadaf.R. Barrett, Richard Kuehn, Jeffery Poole, Steve Scientific Computing Group CSCS-Swiss National Supercomputing Center Manno Switzerland Computer Science and Mathematics Division Oak Ridge National Laboratory Oak Ridge Switzerland

ISBN: (纸本)9780769538020

The building blocks of emerging Petascale massively parallel processing (MPP) systems are multi-core processors with four or more cores as a single processing element and a customized network interface. The resulting memory and communication hierarchy of these platforms are now exposed to application developers and end users by creating a hierarchical or multi-core aware message-passing (MPI) programming interface and by providing a handful of runtime, tunable parameters that allows mapping and control of MPI tasks and message handling. We characterize performance of MPI communication patterns and present strategies for optimizing applications performance on Cray XT series systems that are composed of contemporary AMD processors and a proprietary network infrastructure. We highlight dependencies in its memory and network subsystems, which could influence production-level applications performance. We demonstrate that MPI micro-benchmarks could mislead an application developer or end user since these benchmarks often do not expose the interplay between memory allocation and usage in the user space, which depends on the number of tasks or cores and workload characteristics. Our studies show performance improvements compared to the default options for our target scientific benchmarks and production-level applications. © 2009 IEEE.

关键词： Message passing

来源：评论

学校读者我要写书评

暂无评论

B-SPLINE CNNS ON LIE GROUPS 8

B-SPLINE CNNS ON LIE GROUPS

引用

8th International Conference on Learning Representations, ICLR 2020

作者： Bekkers, Erik J. Amsterdam Machine Learning Lab Informatics Institute University of Amsterdam Netherlands Centre for Analysis and Scientific Computing Applied Mathematics and Computer Science Eindhoven University of Technology Netherlands

Group convolutional neural networks (G-CNNs) can be used to improve classical CNNs by equipping them with the geometric structure of groups. Central in the success of G-CNNs is the lifting of feature maps to higher dimensional disentangled representations in which data characteristics are effectively learned, geometric data-augmentations are made obsolete, and predictable behavior under geometric transformations (equivariance) is guaranteed via group theory. Currently, however, the practical implementations of G-CNNs are limited to either discrete groups (that leave the grid intact) or continuous compact groups such as rotations (that enable the use of Fourier theory). In this paper we lift these limitations and propose a modular framework for the design and implementation of G-CNNs for arbitrary Lie groups. In our approach the differential structure of Lie groups is used to expand convolution kernels in a generic basis of B-splines that is defined on the Lie algebra. This leads to a flexible framework that enables localized, atrous, and deformable convolutions in G-CNNs by means of respectively localized, sparse and non-uniform B-spline expansions. The impact and potential of our approach is studied on two benchmark datasets: cancer detection in histopathology slides in which rotation equivariance plays a key role and facial landmark localization in which scale equivariance is important. In both cases, G-CNN architectures outperform their classical 2D counterparts and the added value of atrous and localized group convolutions is studied in detail. © 2020 8th International Conference on Learning Representations, ICLR 2020. All rights reserved.

关键词： Lie groups

来源：评论

学校读者我要写书评

暂无评论

An efficient geosciences workflow on multi-core processors and GPUs:a case study for aerosol optical depth retrieval from MODIS satellite data

引用

International Journal of Digital Earth 2016年第8期9卷 748-765页

作者： Jia Liu Dustin Feld Yong Xue Jochen Garcke Thomas Soddemann Peiyuan Pan Key Laboratory of Digital Earth Science Institute of Remote Sensing and Digital EarthChinese Academy of SciencesBeijingChina Faculty of Life Sciences and Computing London Metropolitan UniversityLondonUK Fraunhofer Institute for Algorithms and Scientific Computing SCAI Sankt AugustinGermany Institute for Numerical Simulation University of BonnBonnGermany University of Chinese Academy of Sciences BeijingChina Department of Computer Science University of CologneCologneGermany

Quantitative remote sensing retrieval algorithms help understanding the dynamic aspects of Digital ***,the Big Data and complex models in Digital Earth pose grand challenges for computation *** this article,taking the aerosol optical depth(AOD)retrieval as a study case,we exploit parallel computing methods for high efficient geophysical parameter *** present an efficient geocomputation workflow for the AOD calculation from the Moderate Resolution Imaging Spectroradiometer(MODIS)satellite *** to their individual potential for parallelization,several procedures were adapted and implemented for a successful parallel execution on multicore processors and Graphics Processing Units(GPUs).The benchmarks in this paper validate the high parallel performance of the retrieval workflow with speedups of up to 5.x on a multi-core processor with 8 threads and 43.x on a *** specifically address the time-consuming model retrieval part,hybrid parallel patterns which combine the multicore processor’s and the GPU’s compute power were implemented with static and dynamic workload distributions and evaluated on two systems with different CPU–GPU *** is shown that only the dynamic hybrid implementation leads to a greatly enhanced overall exploitation of the heterogeneous hardware environment in varying circumstances.

关键词： Digital earth highperformance computing GPU multi-core hybrid parallel pattern aerosol optical depth retrieval workflow

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：