On partial optimality in multi-label MRFs.
Pushmeet KohliAlexander ShekhovtsovCarsten RotherVladimir KolmogorovPhilip H. S. TorrPublished in: ICML (2008)
Keyphrases
- multi label
- graph cuts
- markov random field
- multi label classification
- text categorization
- image classification
- binary classification
- image annotation
- multi label learning
- belief propagation
- potts model
- higher order
- multi instance
- image segmentation
- text classification
- max margin
- class labels
- multiple labels
- feature selection
- loopy belief propagation
- hierarchical text categorization
- cost sensitive
- maximum a posteriori
- k nearest neighbor
- decision trees
- labeling problems
- computer vision
- protein function prediction
- neural network