Revisiting loss-specific training of filter-based MRFs for image restoration.
Yunjin ChenThomas PockRené RanftlHorst BischofPublished in: CoRR (2014)
Keyphrases
- probabilistic model
- image restoration
- markov random field
- low level vision
- total variation
- edge preserving
- image processing
- regularization parameter
- super resolution
- blurred images
- restoration algorithm
- higher order
- image deblurring
- low level vision problems
- motion blur
- degraded images
- modified hopfield neural network
- iterative image restoration
- kernel estimation
- restored image
- image prior
- high order
- graph cuts
- maximum a posteriori
- adaptive neighborhood
- energy minimization
- total least squares
- gaussian noise
- image deconvolution
- low level vision tasks
- image restoration problems
- blur identification
- image segmentation
- regularization method
- parameter estimation
- training set
- pairwise
- computer vision
- belief propagation
- mrf model
- potential functions