Unsupervised Self-training Algorithm Based on Deep Learning for Optical Aerial Images Change Detection.
Yuan ZhouXiangrui LiPublished in: CoRR (2020)
Keyphrases
- change detection
- aerial images
- deep learning
- training algorithm
- unsupervised learning
- weakly supervised
- back propagation
- neural network
- remote sensing
- remote sensing images
- aerial imagery
- machine learning
- urban areas
- training process
- semi supervised
- satellite images
- support vector machine
- remotely sensed
- remotely sensed images
- supervised learning
- data streams
- rbf neural network
- satellite imagery
- learning algorithm
- supervised classification
- learning rate
- digital elevation models
- image registration
- image analysis
- artificial neural networks
- line segments
- object recognition
- hidden layer
- superpixels
- dimensionality reduction
- probabilistic model
- model selection