Enhancing Semi-Supervised Semantic Segmentation of Remote Sensing Images via Feature Perturbation-Based Consistency Regularization Methods.
Yi XinZide FanXiyu QiYing GengXinming LiPublished in: Sensors (2024)
Keyphrases
- remote sensing images
- semantic segmentation
- regularization methods
- semi supervised
- change detection
- weakly supervised
- multispectral
- remote sensing
- semi supervised learning
- conditional random fields
- scene classification
- superpixels
- object categories
- pairwise
- labeled data
- unlabeled data
- supervised learning
- image features
- computer vision
- unsupervised learning
- higher order
- deformable models
- image data
- object classes
- image analysis
- feature vectors
- active learning
- high resolution
- training set
- high quality
- training data