Graph Segmentation-Based Pseudo-Labeling for Semi-Supervised Pathology Image Classification.
Hong-Kyu ShinKwang-Hyun UhmKyuyeon ChoiZhixin XuSeung-Won JungSung-Jea KoPublished in: IEEE Access (2022)
Keyphrases
- image classification
- semi supervised
- label propagation
- image segmentation
- graph construction
- active learning
- semi supervised learning
- semi supervised classification
- segmentation algorithm
- image labeling
- labeled data
- level set
- unlabeled data
- unsupervised learning
- labeling process
- segmentation method
- region growing
- bag of words
- graph structure
- image features
- feature extraction
- interactive image segmentation
- min cut
- multiscale
- visual words
- image segmentations
- pairwise
- image representation
- directed acyclic graph
- semantic segmentation
- medical images
- graph partitioning
- object segmentation
- multi view
- min cut max flow
- text classification
- shape prior
- pairwise constraints
- directed graph
- graph cuts
- normalized cut
- training data
- partially labeled
- rib cage
- object recognition
- segmented regions
- supervised learning
- co training
- connected components
- conditional random fields
- random walk