Augmenting CRFs with Boltzmann Machine Shape Priors for Image Labeling.
Andrew KaeKihyuk SohnHonglak LeeErik G. Learned-MillerPublished in: CVPR (2013)
Keyphrases
- image labeling
- shape prior
- boltzmann machine
- conditional random fields
- generative model
- bayesian framework
- image segmentation
- crf model
- graph cuts
- prior knowledge
- active contours
- level set
- segmentation method
- deformable models
- shape model
- object segmentation
- probabilistic model
- structured prediction
- global constraints
- markov random field
- graphical models
- energy function
- statistical models
- random fields
- segmentation algorithm
- superpixels
- energy minimization
- hidden markov models
- prior information
- user interaction
- semi supervised
- gaussian process
- markov networks
- higher order
- topic models
- loopy belief propagation
- hidden variables
- information extraction
- semantic segmentation
- neural network