Synthesizing manipulation sequences for under-specified tasks using unrolled Markov Random Fields.
Jaeyong SungBart SelmanAshutosh SaxenaPublished in: IROS (2014)
Keyphrases
- markov random field
- graph cuts
- random fields
- mrf model
- higher order
- parameter estimation
- image segmentation
- potential functions
- energy function
- belief propagation
- maximum a posteriori
- image restoration
- energy minimization
- pairwise
- hidden markov models
- potts model
- loopy belief propagation
- mrf models
- energy minimization problems
- piecewise constant functions
- low level vision
- map inference
- discriminative random fields
- markov networks
- conditional random fields
- map estimation
- iterative conditional
- model selection
- image analysis
- multiscale
- image processing
- machine learning