F.J. Seinstra. User Transparent Parallel Image Processing. PhD thesis, University of Amsterdam, 2003. ISBN 90-5776102 -5. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
User Transparency: A Fully Sequential Programming Model for.. - Seinstra, Koelma (2004) (Correct)
....strict limitations may be imposed on the target platform s size, or the amount of power consumption. In this paper, however, we restrict ourselves to the list as presented here, as this represents the set of general requirements that holds for most image processing research areas. As described in [6], many general purpose parallel architectures (ranging from the Cray family of vector machines, to popular distributed memory multicomputers such as the IBM SP 2, as well as the many shared memory multiprocessors such as the SGI Origin 2000) are potential candidates for high speed execution of ....
....optimization to be performed automatically, fully transparent to the user. This type of optimization is performed by the architecture s scheduling component, described below. A detailed description of the approach of semiempirical modeling, and the applied benchmarking strategy is given in [6,51]. Component 4: database of benchmarking results All benchmarking results are stored on a database of performance values. Although the design and implementation of such a database is of significant importance (especially in case it must be accessed frequently at run time) this topic is too far ....
[Article contains additional citation context not shown here]
Seinstra FJ. User transparent parallel image processing. PhD Thesis, Faculty of Science, University of Amsterdam, The Netherlands, May 2003.
Lazy Parallelization: A Finite State Machine Based.. - Seinstra, Koelma (2003) (2 citations) Self-citation (Seinstra) (Correct)
No context found.
F. Seinstra. User Transparent Parallel Image Processing. PhD thesis, Intelligent Sensory Information Systems, Faculty of Science, University of Amsterdam, May 2003.
Lazy Parallelization: A Finite State Machine Based.. - Seinstra, Koelma (2003) (2 citations) Self-citation (Seinstra) (Correct)
No context found.
F. Seinstra. User Transparent Parallel Image Processing. PhD thesis, Intelligent Sensory Information Systems, Faculty of Science, University of Amsterdam, May 2003.
User Transparency: A Fully Sequential Programming Model for.. - Seinstra, Koelma (2002) Self-citation (Seinstra) (Correct)
No context found.
F.J. Seinstra. User Transparent Parallel Image Processing. PhD thesis, Faculty of Science, University of Amsterdam, The Netherlands, May 2003.
Finite State Machine Based Optimization of Data.. - Seinstra, Koelma.. (2004) Self-citation (Seinstra) (Correct)
No context found.
F.J. Seinstra. User Transparent Parallel Image Processing. PhD thesis, Intelligent Sensory Information Systems, Faculty of Science, University of Amsterdam, The Netherlands, May 2003.
On the Correctness of Lazy Parallelization - Seinstra, Koelma, Bagdanov (2004) Self-citation (Seinstra) (Correct)
No context found.
F.J. Seinstra. User Transparent Parallel Image Processing. PhD thesis, Intelligent Sensory Information Systems, Faculty of Science, University of Amsterdam, The Netherlands, May 2003.
Lazy Parallelization: A Finite State Machine Based.. - Seinstra, Koelma (2003) (2 citations) Self-citation (Seinstra) (Correct)
....time cost estimation, such that the fastest version of the program is represented by the cheapest branch in the graph. Drawback of the extended approach, however, is that it is often costly to actually obtain the cheapest branch. More detailed information related to these issues is presented in [15, 16, 18]. FOR i=0:NrImages 1 DO InputIm = ReadFile( SqrdInputIm = BinPixOp(InputIm, mul , InputIm) FOR j=0:NrSymbols 1 DO IF (i= 0) THEN weights[j] ReadFile( symbols[j] ReadFile( symbols[j] BinPixOp(symbols[j] mul , weights[j] FI FiltIm1 = GenConvOp(SqrdInputIm, ....
....fact that it requires no a priori knowledge regarding the branching behavior of the application at hand. An additional advantage of lazy parallelization is that it requires very little runtime overhead. Also, in our software architecture it proved to be possible to implement the approach elegantly [15]. In conclusion, lazy parallelization on the basis of a finite state machine specification has proven to constitute a surprisingly simple, yet effective method for global optimization of data parallel image processing applications. Essentially, the simplicity stems from the high level ....
F. Seinstra. User Transparent Parallel Image Processing. PhD thesis, Intelligent Sensory Information Systems, Faculty of Science, University of Amsterdam, May 2003.
SmartCam Design Framework - Caarls Jonker And (2003) (Correct)
No context found.
F.J. Seinstra. User Transparent Parallel Image Processing. PhD thesis, University of Amsterdam, 2003. ISBN 90-5776102 -5.
Application Driven Design Of Embedded Real-Time Image.. - Pieter Jonker Wouter (2003) (Correct)
No context found.
F.J. Seinstra, User Transparent Parallel Image Processing, Ph.D. thesis, University of Amsterdam, 2003, ISBN 90-5776-102-5.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC