High-Resolution Fine-Grained Wetland Mapping Based on Class-Balanced Deep Semantic Segmentation Networks.
Yingxin WuYinhe LiuSunan ShiYanfei ZhongPublished in: IGARSS (2023)
Keyphrases
- fine grained
- semantic segmentation
- high resolution
- coarse grained
- object classes
- label transfer
- low resolution
- weakly supervised
- superpixels
- conditional random fields
- scene classification
- access control
- image processing
- super resolution
- high quality
- object categories
- machine learning
- long range
- object class
- probabilistic model
- natural language
- multiscale