Large-Scale Land Cover Mapping with Fine-Grained Classes via Class-Aware Semi-Supervised Semantic Segmentation.
Runmin DongLichao MouMengxuan ChenWeijia LiXin-Yi TongShuai YuanLixian ZhangJuepeng ZhengXiao Xiang ZhuHaohuan FuPublished in: ICCV (2023)
Keyphrases
- fine grained
- semantic segmentation
- land cover
- semi supervised
- object classes
- coarse grained
- class labels
- multispectral
- change detection
- weakly supervised
- satellite images
- remote sensing
- object class
- remote sensing images
- access control
- object categories
- supervised classification
- scene classification
- supervised learning
- conditional random fields
- unlabeled data
- training set
- labeled data
- image features
- pairwise
- metadata
- unsupervised learning
- graph cuts
- object detection
- multi class
- d objects
- superpixels
- image analysis
- data analysis
- object recognition
- high quality
- image processing