Injecting spectral indices to transferable convolutional neural network under imbalanced and noisy labels for Landsat image classification.
Xuemei ZhaoJun WuHaijian WangXingyu GaoLonglong ZhaoPublished in: Int. J. Digit. Earth (2022)
Keyphrases
- convolutional neural network
- image classification
- multispectral images
- multi label
- thematic mapper
- multispectral
- face detection
- bag of words
- image representation
- feature extraction
- hyperspectral
- image features
- pairwise
- supervised classification
- remote sensing images
- hyperspectral imagery
- label noise
- spectral analysis
- segmented images
- class specific
- class imbalance
- neural network
- noisy data
- visual features
- imbalanced data
- satellite images
- satellite imagery
- remotely sensed images
- hyperspectral images
- spatial resolution
- visual words
- image data
- training data
- remote sensing data
- remote sensing
- object detection
- video sequences
- image processing