Mining on Manifolds: Metric Learning Without Labels.
Ahmet IscenGiorgos ToliasYannis AvrithisOndrej ChumPublished in: CVPR (2018)
Keyphrases
- metric learning
- pairwise
- manifold structure
- distance metric learning
- distance metric
- fully supervised
- dimensionality reduction
- feature space
- semi supervised
- machine learning and pattern recognition
- learning tasks
- multi task
- manifold learning
- distance function
- pairwise constraints
- high dimensional
- semi supervised clustering
- low dimensional
- mahalanobis metric
- data mining
- training data
- image classification
- supervised learning
- knowledge discovery
- subject to linear constraints
- background knowledge
- spectral clustering
- multi label
- text mining
- multi class
- face recognition