Convex optimization for multi-class image labeling with a novel family of total variation based regularizers.
Jan LellmannFlorian BeckerChristoph SchnörrPublished in: ICCV (2009)
Keyphrases
- convex optimization
- multi class
- image labeling
- total variation
- structured prediction
- image restoration
- denoising
- image denoising
- conditional random fields
- pairwise
- regularization methods
- low rank
- object detection
- image processing
- support vector machine
- convex optimization problems
- image segmentation
- binary classification
- multi task
- multi class boosting
- feature selection
- natural images
- markov random field
- decision trees
- linear combination
- graph cuts
- text classification
- machine learning