检索结果-内蒙古大学图书馆

Supercomputing Conference

作者： J. Liu V.A. Saletore Department of Computer Science Western Oregon State College Monmouth OR USA Department of Computer Science Oregon State University Corvallis OR USA

The authors present a general approach of self-scheduling a non-uniform parallel loop on a distributed-memory machine. The approach has two phases: a static scheduling phase and a dynamic scheduling phase. In addition to reduce scheduling overhead, using the static scheduling phase allows the data needed by the statically scheduled iterations to be prefetched. The dynamic scheduling phase balances the workload. Data distribution methods for self-scheduling are also the focus of this paper. The authors classify the data distribution methods into four categories and present partial duplication, a method that allows the problem size to grow linearly in the number of processors. The experiments conducted on a 64-node NCUBE show that as much as 79% improvement is achieved over static scheduling on the generation of a false-color image.

关键词： Processor scheduling Dynamic scheduling Application software Computer science Concurrent computing Educational institutions Prefetching Focusing Acceleration parallel processing

来源：评论

学校读者我要写书评

暂无评论

Testing parallel evolution strategies on the quadratic assignment problem

Testing parallel evolution strategies on the quadratic assig...

引用

IEEE International Conference on Systems, Man and Cybernetics

作者： I. de Falco R. del Balio E. Tarantino Istituto per la Ricerca sui Sistemi Informatici Paralleli (IRSIP) CNR Naples Italy

parallel evolution strategies are demonstrating to be worthwhile in a variety of contexts. In this paper, besides the classical genetic and evolutionary strategies, a hybrid evolutionary approach which incorporates memory of the search history within the structure is analyzed. The parallel evolution algorithms are mapped on a distributed memory MIMD multicomputer whose processors are configured in a torus topology. The simulations are conducted using the quadratic assignment problem as an artificial environment. The relationship between genetic representations and recombination operators is investigated. The experimental results obtained show the value of structures richer than bit strings and the effectiveness of memory for the evolution process.< >

关键词： Testing Evolutionary computation Optimization methods History Topology Genetic algorithms Delay Neural networks image processing Machine learning

来源：评论

学校读者我要写书评

暂无评论

THE PARTITIONING PROBLEM FOR A CLASS OF DATA parallel ALGORITHMS

引用

parallel COMPUTING 1992年第8期18卷 867-878页

作者： THUNE, M Department of Scientific Computing Uppsala University Box 120 S-751 04 Uppsala Sweden

A solution to the partitioning problem is presented for a class of data parallel algorithms (including for example explicit difference methods for time-dependent PDE, and image processing algorithms based on local filters). Conditions are formulated, that characterize the optimal partitioning. From them, an explicit formula for the optimal partitioning is derived, which is valid in special cases. For the general case, the conditions provide a basis for the formulation of iterative partitioning algorithms. One such algorithm is proposed. The partitioning algorithm is intended as a tool to be used in utility routines or, ultimately, compilers, to enhance SPMD programming of MIMD-type computers with distributed memory. Results from an application in image analysis show that the algorithm is suitable for this purpose.

关键词： PARTITIONING DATA parallelISM MIMD COMPUTERS distributed MEMORY SPMD PROGRAMMING image ANALYSIS EXPLICIT DIFFERENCE methods

来源：评论

学校读者我要写书评

暂无评论

processing OF ACOUSTIC-SIGNALS IN A COCHLEAR MODEL INCORPORATING LATERALLY COUPLED SUPPRESSIVE ELEMENTS

引用

NEURAL NETWORKS 1992年第1期5卷 19-34页

作者： DENG, L University of Waterloo Canada

Development of hearing machines with brain-like performance can be approached by faithfully modeling the human auditory nervous system. As an initial step in this effort, a composite cochlear model has been built which processes acoustic signals in a manner consistent with physiological responses recorded from the mammalian auditory nerve. The model integrates a current state of knowledge about cochlear function and provides a coherent picture of the representation of acoustic signals in the pattern of neural impulses distributed across a tonotopically organized array of auditory-nerve fibers. The model explicitly represents the active electro-mechanical responses of outer hair cells as a possible mechanism for adaptive control of basilar membrane damping, and introduces lateral coupling of resistive elements in the cochlear mechanics based on this mechanism. The output of the model is represented as an acoustic image being transferred to the central auditory nervous system in a massively parallel fashion over the auditory nerve. The impact of the cochlear nonlinearity on this image is explored. Experiments on applying this model to speech analysis show advantages of the cochlear modeling approach over traditional linear analysis methods.

关键词： COCHLEAR MODEL SPEECH processing NEURAL RESPONSES HEARING MACHINES DYNAMIC NONLINEARITY ADAPTIVE GAIN CONTROL LATERAL SUPPRESSION OUTER HAIR CELLS

来源：评论

学校读者我要写书评

暂无评论

Proceedings of the Scalable High Performance Computing Conference-SHPCC-92

Proceedings of the Scalable High Performance Computing Confe...

引用

Proccedings of the Scalable High Performance Computing Conference-SHPCC-92

ISBN: (纸本)0818627751

This conference proceedings contains 70 papers. The topics discussed are image processing applications;performance analysis of parallel and multiprocessing programs: applications of parallel architectures to simulation and real time control;parallel languages;network attached storage systems;parallel algorithms;systems support for languages;molecular dynamics;software tools;object oriented programming;computational fluid dynamics;parallel computer architectures;load balancing;computational methods;interprocessor communication;distributed memory multicomputers;compilers;conversions of sequential to parallel programs;parallel program applications;debugging;hypercubes;irregular problems;and systems issues.

关键词： parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

parallel distributed algorithm for texture boundary localization

Parallel distributed algorithm for texture boundary localiza...

引用

Neural and Stochastic methods in image and Signal processing

作者： Yhann, Stephan D. Young, Tzay S. Scripps Inst. of Oceanography Coral Gables FL USA

ISBN: (纸本)0819409391

In most instances the boundaries between textured regions are defined by the gray level contrasts which result from the local interaction between the texture elements in each region. In such cases, the boundaries can be accurately characterized by gray level edge segments. Using these edge segments to localize the texture boundary directly addresses the major problem associated with texture segmentation, namely the localization verses classification accuracy conflict. The accuracy of segmentation methods which rely only on spatially distributed properties to characterize the texture, is limited to the spacial extent of the property used. In contrast, gray level edges are significantly more localized. However, before they can be of any use, the gray level edge segments defining the texture boundary must be isolated from the edges defining the texture elements. In this paper, we define a set of properties to do this. We also incorporate these properties into a parallel distributed algorithm which is used to segment a set of sample texture images.

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

image filtering on slotted ring multiple DSP systems

Image filtering on slotted ring multiple DSP systems

引用

作者： Oakley, D. Bruce Carleton University (Canada)

学位级别：M.Eng.

This thesis describes the partitioning of the linear image filtering problem for multiple DSP systems interconnected by slotted ring networks. An analysis of various partitioning methods is presented. The block parallel transform method, based on overlap-and-save processing in the frequency domain, is identified as an effective algorithm for image filtering with multiple processors. In this algorithm, image blocks are distributed to separate processors where filtering operations occur. The slotted ring architecture is shown to effectively support the block parallel transform method. Other architectures are identified which can provide comparable performance. An implementation of the block parallel transform method on a representative slotted ring system is described. processing times correspond well with the analytical model and speed increases are roughly proportional to the number of processors used. A system with four processors filtered a 128 x 128 image 6 to 7 times faster than PC-based software.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The odd-even expansion storage scheme and its implementation issues

The odd-even expansion storage scheme and its implementation...

引用

International Symposium on parallel processing

作者： Z. Liu J.-H. You X. Li Department of Computing Science University of Alberta Edmonton AB Canada

The authors present a parallel storage scheme to distribute the elements of an N*N matrix over N memory banks, where N is any (odd or even) power of two, such that any rows, columns, forward and backward diagonals, and square or rectangular blocks can be accessed simultaneously without memory conflict. They present a simple scheme for address generation, which requires only logic operations and can be completed in constant time. They present two network implementation methods for data alignments for this storage scheme. Different from previously proposed routing algorithms, the algorithms for hypercube routing in this paper are free from network conflict. They do not require buffering and time length of a 'step' is shorter, therefore they are more efficient in terms of both hardware cost and speed. The authors also present a simple MIN implementation scheme for the realization of the data alignments. Schemes for processing smaller matrices efficiently on larger scale systems are also developed.< >

关键词： Routing Hypercubes image processing Concurrent computing distributed computing Logic Hardware Costs parallel processing Multiprocessor interconnection networks

来源：评论

学校读者我要写书评

暂无评论

Study of affine image warping on a linear processing array

Study of affine image warping on a linear processing array

引用

International Conference on Pattern Recognition

作者： G. Gusciora J.A. Webb Electrical and Computer Engineering Department Carnegie Mellon University Pittsburgh PA USA School of Computer Science Carnegie Mellon University Pittsburgh PA USA

Several methods for parallel affine image warping on a linear processor array are considered. The methods were implemented on the Carnegie Mellon Warp machine and the Carnegie Mellon-Intel Corporation iWarp computer (treated as a linear array), and performance figures are provided. Both systolic methods, which feed one of the images in a stream, and non-systolic methods, which partition both images, are treated. A scanline method that combines some of the features of both, but which requires a fast transposed method is also described. The authors articulate three characteristics that affect the design of parallel image warping algorithms: affine warping is easily invertible, the mapping is known at the start of execution, and nearby input pixels map to nearby output pixels. The authors conclude that non-systolic methods give slightly better execution time and are easier to programs than systolic methods but require much larger processor memories.< >

关键词： Interpolation Forward contracts Lagrangian functions Polynomials Pixel Runtime distributed computing parallel machines Concurrent computing Feeds

来源：评论

学校读者我要写书评

暂无评论

DYNAMIC INTELLIGENT SCHEDULING AND CONTROL OF RECONFIGURABLE parallel ARCHITECTURES FOR COMPUTER VISION image-processing

引用

JOURNAL OF parallel AND distributed COMPUTING 1991年第3期13卷 273-285页

作者： WEIL, F JAMIESON, LH DELP, EJ PURDUE UNIV SCH ELECT ENGNW LAFAYETTEIN 47907

This paper presents a system for dynamic intelligent scheduling and control of reconfigurable parallel processors. The purpose of the system is to provide a rapid prototyping capability for computer vision/image processing tasks. The scheduler particularly addresses the problems of algorithms with execution times that depend on the image data and processing scenarios that vary dynamically based on the input image. Since conventional scheduling methods cannot produce schedules for most tasks of this type, a dynamic controller is used to schedule the task and reconfigure the machine “on the fly.” This dynamic scheduling system attempts to balance the overall processing scenario with the needs of the individual routines that make up the task. This paper discusses the implementation of the DISC ( D ynamic I ntelligent S cheduling and C ontrol) system. Emphasis is on the scheduling heuristics as they apply to a reconfigurable parallel processor, the information in the system database, and the use of the system for prototyping computer vision/image processing tasks on a partitionable parallel system.

关键词： Computer Vision

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：