Exploiting Multi-Scale Parallel Self-Attention and Local Variation via Dual-Branch Transformer-CNN Structure for Face Super-Resolution.
Jingang ShiYusi WangZitong YuGuanxin LiXiaopeng HongFei WangYihong GongPublished in: IEEE Trans. Multim. (2024)
Keyphrases
- super resolution
- low resolution
- high resolution
- multiscale
- image reconstruction
- image restoration
- motion estimation
- image resolution
- low resolution images
- super resolution reconstruction
- image super resolution
- motion blur
- single frame
- multi frame
- resolution enhancement
- higher resolution
- adaptive regularization
- high quality
- light field
- high resolution images
- depth map
- scale space
- super resolved
- video super resolution
- image processing
- visual attention
- principal component analysis
- single image super resolution
- face hallucination
- facial expressions
- super resolution image reconstruction
- low resolution face images