Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Parallel and Distributed Vision Algorithms Using Dual Decomposition

Strandmark, Petter LU ; Kahl, Fredrik LU and Schoenemann, Thomas LU (2011) In Computer Vision and Image Understanding 115(12). p.1721-1732
Abstract
We investigate dual decomposition approaches for optimization problems arising in low-level vision. Dual decomposition can be used to parallelize existing algorithms, reduce memory requirements and to obtain approximate solutions of hard problems. An extensive set of experiments are performed for a variety of application problems including graph cut segmentation, curvature regularization and more generally the optimization of MRFs. We demonstrate that the technique can be useful for desktop computers, graphical processing units and supercomputer clusters. To facilitate further research, an implementation of the decomposition methods is made publicly available.
Please use this url to cite or link to this publication:
author
; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Graph cuts, Dual decomposition, Parallel, MRF, MPI, GPU
in
Computer Vision and Image Understanding
volume
115
issue
12
pages
1721 - 1732
publisher
Elsevier
external identifiers
  • wos:000297085500011
  • scopus:80455164637
ISSN
1077-3142
DOI
10.1016/j.cviu.2011.06.012
language
English
LU publication?
yes
id
bd4eda0b-f116-4a63-bafa-a6d12c4f3cde (old id 2060506)
alternative location
http://www.sciencedirect.com/science/article/pii/S1077314211001652
date added to LUP
2016-04-01 11:15:35
date last changed
2022-03-12 21:06:08
@article{bd4eda0b-f116-4a63-bafa-a6d12c4f3cde,
  abstract     = {{We investigate dual decomposition approaches for optimization problems arising in low-level vision. Dual decomposition can be used to parallelize existing algorithms, reduce memory requirements and to obtain approximate solutions of hard problems. An extensive set of experiments are performed for a variety of application problems including graph cut segmentation, curvature regularization and more generally the optimization of MRFs. We demonstrate that the technique can be useful for desktop computers, graphical processing units and supercomputer clusters. To facilitate further research, an implementation of the decomposition methods is made publicly available.}},
  author       = {{Strandmark, Petter and Kahl, Fredrik and Schoenemann, Thomas}},
  issn         = {{1077-3142}},
  keywords     = {{Graph cuts; Dual decomposition; Parallel; MRF; MPI; GPU}},
  language     = {{eng}},
  number       = {{12}},
  pages        = {{1721--1732}},
  publisher    = {{Elsevier}},
  series       = {{Computer Vision and Image Understanding}},
  title        = {{Parallel and Distributed Vision Algorithms Using Dual Decomposition}},
  url          = {{http://dx.doi.org/10.1016/j.cviu.2011.06.012}},
  doi          = {{10.1016/j.cviu.2011.06.012}},
  volume       = {{115}},
  year         = {{2011}},
}