D-ESRGAN: A Dual-Encoder GAN with Residual CNN and Vision Transformer for Iris Image Super-Resolution.
Caiyong WangTianhao LuGaosheng WuYunlong WangZhenan SunPublished in: IJCB (2022)
Keyphrases
- image super resolution
- dictionary learning
- low resolution
- super resolution
- computer vision
- kernel regression
- edge preserving
- image registration
- high resolution
- low resolution images
- image patches
- motion estimation
- image processing
- high resolution images
- map estimation
- sparse representation
- bit rate
- maximum a posteriori
- sparse coding
- image restoration
- class specific
- data sets
- facial images
- image reconstruction
- higher order
- least squares