Semi-supervised Multiclass Clustering Based on Signed Total Variation.
Peter BergerThomas DittrichGabor HannakGerald MatzPublished in: ICASSP (2019)
Keyphrases
- multi class
- total variation
- semi supervised
- pairwise
- image restoration
- image denoising
- denoising
- semi supervised learning
- convex optimization
- regularization term
- support vector machine
- labeled data
- image processing
- active learning
- unlabeled data
- binary classification
- supervised learning
- feature selection
- pairwise constraints
- regularization methods
- cost sensitive
- low rank
- gaussian process
- binary classifiers
- multiclass classification
- computer vision
- multi task
- metric learning
- markov random field
- machine learning
- loss function
- multi label classification
- training data
- regularization parameter
- super resolution
- natural images
- learning problems