A balanced random learning strategy for CNN based Landsat image segmentation under imbalanced and noisy labels.
Xuemei ZhaoYong ChengLuo LiangHaijian WangXingyu GaoJun WuPublished in: Pattern Recognit. (2023)
Keyphrases
- learning strategies
- image segmentation
- segmented images
- active learning
- label noise
- online learning
- cellular neural networks
- multispectral
- imbalanced class distribution
- blended learning
- region growing
- graph cuts
- image processing
- learning capabilities
- training data
- markov random field
- image analysis
- active contours
- satellite images
- supervised classification
- computer vision
- computer based instruction
- class imbalance
- professional development
- training set
- imbalanced data
- convolutional neural network
- training examples
- human learning
- satellite imagery
- pairwise
- virtual learning environments
- class distribution
- fuzzy c means
- level set
- class labels
- segmentation algorithm