English

Submodularization for Quadratic Pseudo-Boolean Optimization

Computer Vision and Pattern Recognition 2014-04-17 v2

Abstract

Many computer vision problems require optimization of binary non-submodular energies. We propose a general optimization framework based on local submodular approximations (LSA). Unlike standard LP relaxation methods that linearize the whole energy globally, our approach iteratively approximates the energies locally. On the other hand, unlike standard local optimization methods (e.g. gradient descent or projection techniques) we use non-linear submodular approximations and optimize them without leaving the domain of integer solutions. We discuss two specific LSA algorithms based on "trust region" and "auxiliary function" principles, LSA-TR and LSA-AUX. These methods obtain state-of-the-art results on a wide range of applications outperforming many standard techniques such as LBP, QPBO, and TRWS. While our paper is focused on pairwise energies, our ideas extend to higher-order problems. The code is available online (http://vision.csd.uwo.ca/code/).

Keywords

Cite

@article{arxiv.1311.1856,
  title  = {Submodularization for Quadratic Pseudo-Boolean Optimization},
  author = {Lena Gorelick and Yuri Boykov and Olga Veksler and Ismail Ben Ayed and Andrew Delong},
  journal= {arXiv preprint arXiv:1311.1856},
  year   = {2014}
}

Comments

8 pages, 5 figures, to appear at IEEE conference on Computer Vision and Pattern Recognition (CVPR), June 2014

R2 v1 2026-06-22T02:03:26.537Z