Adjustable super-resolution network via deep supervised learning and progressive self-distillation.
Juncheng LiFaming FangTieyong ZengGuixu ZhangXizhao WangPublished in: Neurocomputing (2022)
Keyphrases
- super resolution
- low resolution
- high resolution
- supervised learning
- image reconstruction
- motion estimation
- low resolution images
- resolution enhancement
- image super resolution
- image restoration
- image resolution
- single frame
- high resolution images
- super resolution reconstruction
- single image super resolution
- super resolved
- motion blur
- depth map
- multi frame
- high quality
- adaptive regularization
- video super resolution
- higher resolution
- light field
- unsupervised learning
- active learning
- image patches
- training data
- degraded images
- machine learning