Protocon: Pseudo-Label Refinement via Online Clustering and Prototypical Consistency for Efficient Semi-Supervised Learning.
Islam NassarMunawar HayatEhsan AbbasnejadHamid RezatofighiGholamreza HaffariPublished in: CVPR (2023)
Keyphrases
- semi supervised learning
- semi supervised
- label propagation
- unsupervised learning
- unlabeled data
- labeled data
- semi supervised clustering
- semi supervised classification
- multi label classification
- manifold regularization
- machine learning
- dirichlet process mixture models
- text categorization
- graph based semi supervised learning
- active learning
- supervised learning
- clustering method
- co training
- data points
- graph construction
- clustering algorithm
- class labels
- training data
- partially labeled
- multi instance learning
- document clustering
- transfer learning
- pair wise constraints
- unlabeled instances
- data sets
- feature extraction
- k means
- labeled and unlabeled data
- unsupervised clustering
- label information
- spectral clustering