Task-Driven Learned Hyperspectral Data Reduction Using End-to-End Supervised Deep Learning.
Mathé T. ZeegersDaniël Maria PeltTristan van LeeuwenRobert van LiereKees Joost BatenburgPublished in: J. Imaging (2020)
Keyphrases
- end to end
- deep learning
- hyperspectral
- data reduction
- unsupervised learning
- feature selection
- weakly supervised
- hyperspectral images
- remote sensing
- multispectral
- data compression
- machine learning
- infrared
- image data
- model selection
- preprocessing
- knowledge discovery
- spectral bands
- data analysis
- information content
- dimensionality reduction
- classification rules
- supervised learning
- semi supervised
- singular value decomposition
- learning algorithm
- classification accuracy
- data mining
- high dimensionality
- data quality
- clustering algorithm
- databases
- compression algorithm
- topic models
- bit rate
- expectation maximization