Parameter Estimation and Energy Minimization for Region-Based Semantic Segmentation.
M. Pawan KumarHaithem TurkiDan PrestonDaphne KollerPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2015)
Keyphrases
- parameter estimation
- energy minimization
- semantic segmentation
- markov random field
- conditional random fields
- image segmentation
- pixel labeling
- energy function
- graph cuts
- random fields
- superpixels
- belief propagation
- higher order
- long range
- expectation maximization
- statistical models
- scene classification
- image understanding
- em algorithm
- pairwise
- segmentation algorithm
- maximum a posteriori
- multiscale
- segmentation method
- weakly supervised
- least squares
- object categories
- maximum likelihood
- object classes
- probabilistic relaxation
- model selection
- computer vision
- object class
- image features
- pascal voc
- object recognition
- deformable models
- mean shift
- probabilistic model