DMCA
MAP-Inference on Large Scale Higher-Order Discrete Graphical Models by Fusion Moves
Cached
Download Links
| Citations: | 2 - 1 self |
BibTeX
@MISC{Beier_map-inferenceon,
author = {Thorsten Beier},
title = {MAP-Inference on Large Scale Higher-Order Discrete Graphical Models by Fusion Moves},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Many computer vision problems can be cast into optimization prob-lems over discrete graphical models also known as Markov or conditional random fields. Standard methods are able to solve those problems quite efficiently. However, problems with huge label spaces and or higher-order structure remain challenging or intractable even for approximate methods. We reconsider the work of Lempitsky et al. 2010 on fusion moves and apply it to general discrete graphical models. We propose two alternatives for calculating fusion moves that outperform the standard in several appli-cations. Our generic software framework allows us to easily use different proposal generators which spans a large class of inference algorithms and thus makes exhaustive evaluation feasible. Because these fusion algorithms can be applied to models with huge label spaces and higher-order terms, they might stimulate and support research of such models which may have not been possible so far due to the lack of adequate inference methods. 1
Keyphrases
fusion move huge label space different proposal generator optimization prob-lems higher-order structure exhaustive evaluation feasible conditional random field general discrete graphical model generic software framework fusion algorithm standard method adequate inference method many computer vision problem discrete graphical model large class approximate method inference algorithm several appli-cations higher-order term
