A class-specific metric learning approach for graph embedding by information granulation.
Luca BaldiniAlessio MartinoAntonello RizziPublished in: Appl. Soft Comput. (2022)
Keyphrases
- class specific
- metric learning
- graph embedding
- semi supervised
- dimensionality reduction
- multiple features
- dictionary learning
- image classification
- sparse representation
- semi supervised learning
- object class
- low dimensional
- labeled data
- pairwise
- distance metric
- principal component analysis
- object categories
- sparse coding
- unlabeled data
- feature space
- active learning
- high dimensional
- pattern recognition
- distance function
- unsupervised learning
- high dimensional data
- learning tasks
- generative model
- supervised learning
- feature extraction
- multi task
- euclidean distance
- object classes
- feature selection
- multiscale
- object detection