Solving Markov Random Fields with Spectral Relaxation

Timothee Cour, Jianbo Shi
; Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics, PMLR 2:75-82, 2007.

Abstract

Markov Random Fields (MRFs) are used in a large array of computer vision and maching learning applications. Finding the Maximum Aposteriori (MAP) solution of an MRF is in general intractable, and one has to resort to approximate solutions, such as Belief Propagation, Graph Cuts, or more recently, approaches based on quadratic programming. We propose a novel type of approximation, Spectral relaxation to Quadratic Programming (SQP). We show our method offers tighter bounds than recently published work, while at the same time being computationally efficient. We compare our method to other algorithms on random MRFs in various settings.

Cite this Paper


BibTeX
@InProceedings{pmlr-v2-cour07a, title = {Solving Markov Random Fields with Spectral Relaxation}, author = {Timothee Cour and Jianbo Shi}, booktitle = {Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics}, pages = {75--82}, year = {2007}, editor = {Marina Meila and Xiaotong Shen}, volume = {2}, series = {Proceedings of Machine Learning Research}, address = {San Juan, Puerto Rico}, month = {21--24 Mar}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v2/cour07a/cour07a.pdf}, url = {http://proceedings.mlr.press/v2/cour07a.html}, abstract = {Markov Random Fields (MRFs) are used in a large array of computer vision and maching learning applications. Finding the Maximum Aposteriori (MAP) solution of an MRF is in general intractable, and one has to resort to approximate solutions, such as Belief Propagation, Graph Cuts, or more recently, approaches based on quadratic programming. We propose a novel type of approximation, Spectral relaxation to Quadratic Programming (SQP). We show our method offers tighter bounds than recently published work, while at the same time being computationally efficient. We compare our method to other algorithms on random MRFs in various settings.} }
Endnote
%0 Conference Paper %T Solving Markov Random Fields with Spectral Relaxation %A Timothee Cour %A Jianbo Shi %B Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2007 %E Marina Meila %E Xiaotong Shen %F pmlr-v2-cour07a %I PMLR %J Proceedings of Machine Learning Research %P 75--82 %U http://proceedings.mlr.press %V 2 %W PMLR %X Markov Random Fields (MRFs) are used in a large array of computer vision and maching learning applications. Finding the Maximum Aposteriori (MAP) solution of an MRF is in general intractable, and one has to resort to approximate solutions, such as Belief Propagation, Graph Cuts, or more recently, approaches based on quadratic programming. We propose a novel type of approximation, Spectral relaxation to Quadratic Programming (SQP). We show our method offers tighter bounds than recently published work, while at the same time being computationally efficient. We compare our method to other algorithms on random MRFs in various settings.
RIS
TY - CPAPER TI - Solving Markov Random Fields with Spectral Relaxation AU - Timothee Cour AU - Jianbo Shi BT - Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics PY - 2007/03/11 DA - 2007/03/11 ED - Marina Meila ED - Xiaotong Shen ID - pmlr-v2-cour07a PB - PMLR SP - 75 DP - PMLR EP - 82 L1 - http://proceedings.mlr.press/v2/cour07a/cour07a.pdf UR - http://proceedings.mlr.press/v2/cour07a.html AB - Markov Random Fields (MRFs) are used in a large array of computer vision and maching learning applications. Finding the Maximum Aposteriori (MAP) solution of an MRF is in general intractable, and one has to resort to approximate solutions, such as Belief Propagation, Graph Cuts, or more recently, approaches based on quadratic programming. We propose a novel type of approximation, Spectral relaxation to Quadratic Programming (SQP). We show our method offers tighter bounds than recently published work, while at the same time being computationally efficient. We compare our method to other algorithms on random MRFs in various settings. ER -
APA
Cour, T. & Shi, J.. (2007). Solving Markov Random Fields with Spectral Relaxation. Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics, in PMLR 2:75-82

Related Material