Closed-Form Training of Conditional Random Fields for Large Scale Image Segmentation.
Alexander KolesnikovMatthieu GuillauminVittorio FerrariChristoph H. LampertPublished in: CoRR (2014)
Keyphrases
- closed form
- conditional random fields
- markov random field
- image segmentation
- structured prediction
- random fields
- sequence labeling
- segmentation method
- probabilistic model
- partition function
- error correcting output coding
- graphical models
- hidden markov models
- higher order
- information extraction
- generative model
- named entity recognition
- maximum entropy
- web page prediction
- graph cuts
- weighted sums
- crf model
- multiscale
- pairwise
- approximate inference
- parameter learning
- training set
- training samples
- superpixels
- markov networks
- loopy belief propagation
- closed form solutions
- cost function
- active contours
- probabilistic graphical models
- object segmentation
- segmentation algorithm
- energy function
- energy minimization
- machine learning