Getting pixels and regions to agree with conditional random fields.
Devis TuiaMichele VolpiGabriele MoserPublished in: IGARSS (2016)
Keyphrases
- conditional random fields
- homogeneous regions
- input image
- image pixels
- hidden markov models
- probabilistic model
- graphical models
- random fields
- image labeling
- sequence labeling
- higher order
- information extraction
- image regions
- generative model
- crf model
- markov random field
- parameter learning
- structured prediction
- maximum entropy
- structured learning
- markov models
- web page prediction
- semi markov
- named entity recognition
- superpixels
- pairwise
- protein fold recognition
- markov networks
- belief propagation
- feature selection