Learnable Manifold Alignment (LeMA) : A Semi-supervised Cross-modality Learning Framework for Land Cover and Land Use Classification.
Danfeng HongNaoto YokoyaNan GeJocelyn ChanussotXiao Xiang ZhuPublished in: CoRR (2019)
Keyphrases
- land cover
- supervised learning
- supervised classification
- semi supervised
- land cover classification
- labeled and unlabeled data
- remote sensing
- learning algorithm
- multispectral
- learning tasks
- unsupervised learning
- remote sensing images
- learning process
- active learning
- image classification
- change detection
- semi supervised learning
- unlabeled data
- reinforcement learning
- data mining
- machine learning
- machine learning algorithms
- pattern recognition
- decision trees
- feature selection
- model selection
- prior knowledge
- feature vectors
- co training
- pairwise
- manifold alignment