PolSAR Image Semantic Segmentation Based on Deep Transfer Learning - Realizing Smooth Classification With Small Training Sets.
Wenjin WuHailei LiXinwu LiHuadong GuoLu ZhangPublished in: IEEE Geosci. Remote. Sens. Lett. (2019)
Keyphrases
- semantic segmentation
- transfer learning
- scene classification
- image classification
- conditional random fields
- text classification
- superpixels
- input image
- image features
- machine learning
- image content
- multiscale
- object recognition
- image representation
- feature extraction
- decision trees
- image collections
- unsupervised learning
- feature space
- reinforcement learning
- data sets
- active learning
- visual features
- segmentation method
- image retrieval
- support vector
- bounding box
- feature vectors
- test images
- object categories
- feature selection
- image segmentation