Accelerated Inference in Markov Random Fields via Smooth Riemannian Optimization.
Siyi HuLuca CarlonePublished in: IEEE Robotics Autom. Lett. (2019)
Keyphrases
- markov random field
- dual decomposition
- random fields
- map inference
- loopy belief propagation
- efficient inference
- energy minimization
- higher order
- parameter estimation
- belief propagation
- mrf model
- graph cuts
- piecewise constant functions
- energy function
- conditional random fields
- partition function
- smoothness prior
- maximum a posteriori
- image segmentation
- pairwise
- message passing
- undirected graphical models
- image restoration
- image restoration and reconstruction
- bayesian networks
- low level vision
- parameter learning
- probabilistic inference
- graphical models
- factor graphs
- evolutionary algorithm
- map estimation
- potential functions
- markov chain monte carlo
- potts model
- iterative conditional
- higher order cliques
- exact inference
- textured images
- probability distribution
- non stationary
- generalized belief propagation
- posterior distribution
- computer vision