Non-homogeneous Conditional Random Fields for Contextual Image Segmentation.
Olfa BesbesNozha BoujemaaZiad BelhadjPublished in: ISM (2008)
Keyphrases
- conditional random fields
- image segmentation
- markov random field
- contextual features
- segmentation method
- random fields
- contextual information
- graphical models
- graph cuts
- long range
- probabilistic model
- sequence labeling
- information extraction
- maximum entropy
- higher order
- hidden markov models
- crf model
- image labeling
- pairwise
- structured learning
- structured prediction
- markov models
- generative model
- energy minimization
- multiscale
- superpixels
- belief propagation
- named entity recognition
- computer vision
- loopy belief propagation
- image processing
- approximate inference
- parameter learning
- probabilistic graphical models
- parameter estimation
- undirected graphical models
- semi markov
- expectation maximization
- shape prior
- markov networks
- energy function
- web documents
- segmentation algorithm