FBRNN: feedback recurrent neural network for extreme image super-resolution.
Junyeop LeeJaihyun ParkKanghyu LeeJeongki MinGwantae KimBokyeung LeeBonhwa KuDavid K. HanHanseok KoPublished in: CVPR Workshops (2020)
Keyphrases
- recurrent neural networks
- image super resolution
- echo state networks
- super resolution
- low resolution
- dictionary learning
- complex valued
- neural network
- recurrent networks
- feed forward
- image registration
- maximum a posteriori
- low resolution images
- map estimation
- edge preserving
- sparse representation
- kernel regression
- artificial neural networks
- high resolution
- image patches
- sparse coding
- fuzzy logic
- input image
- image reconstruction
- high resolution images
- maximum likelihood
- feature space
- object recognition
- image segmentation
- computer vision