A Deep Learning Approach to Block-based Compressed Sensing of Images.
Amir AdlerDavid BoublilMichael EladMichael ZibulevskyPublished in: CoRR (2016)
Keyphrases
- deep learning
- compressed sensing
- input image
- image features
- image reconstruction
- random projections
- image classification
- image retrieval
- fourier domain
- multiple images
- natural images
- test images
- segmentation algorithm
- unsupervised learning
- similarity measure
- training data
- signal processing
- motion estimation
- weakly supervised
- active learning
- multiscale