Fast Approximate Energy Minimization via Graph Cuts
IEEE Transactions on Pattern Analysis and Machine Intelligence
Segmentation by Grouping Junctions
CVPR '98 Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Convergent Tree-Reweighted Message Passing for Energy Minimization
IEEE Transactions on Pattern Analysis and Machine Intelligence
A Linear Programming Approach to Max-Sum Problem: A Review
IEEE Transactions on Pattern Analysis and Machine Intelligence
Exact solution of permuted submodular minsum problems
EMMCVPR'07 Proceedings of the 6th international conference on Energy minimization methods in computer vision and pattern recognition
Stop condition for subgradient minimization in dual relaxed (max,+) problem
EMMCVPR'11 Proceedings of the 8th international conference on Energy minimization methods in computer vision and pattern recognition
| Hi-index | 0.00 |
A formal analysis of so-called diffusion algorithms is performed. They are frequently used in structural recognition but are rather poorly theoretically studied. These algorithms are analyzed from the viewpoint of their ability to optimize a function of many discrete variables, which is represented as the sum of many terms each of which depends on only two variables. It is proved that, under some stop condition, a diffusion algorithm approximately solves certain subclasses of optimization problems with any predefined nonzero error. The totality of problems solved by diffusion algorithms includes all so-called acyclic and supermodular optimization problems and also some other problems for which solution algorithms are unknown.