DSegAN: A Deep Light-weight Segmentation-based Attention Network for Image Restoration.
Alireza EsmaeilzehiM. Omair AhmadM. N. S. SwamyPublished in: ISCAS (2022)
Keyphrases
- image restoration
- lightweight
- markov random field models
- low level vision
- edge preserving
- total variation
- communication infrastructure
- super resolution
- wireless sensor networks
- image processing
- image deblurring
- restoration method
- multiscale
- blind image restoration
- blurred images
- modified hopfield neural network
- iterative image restoration
- blur identification
- regularization method
- image segmentation
- markov random field
- regularization parameter
- gauss markov
- level set
- segmentation method
- segmentation algorithm
- energy function
- network structure
- mumford shah
- point spread function
- restoration algorithm
- degraded images
- energy functional
- image restoration problems
- gaussian noise
- kernel estimation
- region growing
- motion estimation