Tree-based iterated local search for Markov random fields with applications in image analysis
Data(s) |
01/01/2014
|
---|---|
Resumo |
The maximum a posteriori assignment for general structure Markov random fields is computationally intractable. In this paper, we exploit tree-based methods to efficiently address this problem. Our novel method, named Tree-based Iterated Local Search (T-ILS), takes advantage of the tractability of tree-structures embedded within MRFs to derive strong local search in an ILS framework. The method efficiently explores exponentially large neighborhoods using a limited memory without any requirement on the cost functions. We evaluate the T-ILS on a simulated Ising model and two real-world vision problems: stereo matching and image denoising. Experimental results demonstrate that our methods are competitive against state-of-the-art rivals with significant computational gain. |
Identificador | |
Idioma(s) |
eng |
Publicador |
Springer |
Relação |
http://dro.deakin.edu.au/eserv/DU:30073006/tran-treebasediterated-2014.pdf http://www.dx.doi.org/10.1007/s10732-014-9270-1 |
Direitos |
2014, Springer |
Palavras-Chave | #Belief propagation #Iterated local search #MAP assignment #Markov random fields #Strong local search #Science & Technology #Technology #Computer Science, Artificial Intelligence #Computer Science, Theory & Methods #Computer Science #BELIEF-PROPAGATION #ENERGY MINIMIZATION #STATISTICAL-ANALYSIS #GENETIC ALGORITHM #MAP ESTIMATION #GRAPH CUTS #SEGMENTATION #OPTIMIZATION #MRF #DISTRIBUTIONS |
Tipo |
Journal Article |