Table 1 Classical algorithms for image super-resolution and their characteristics.
Methods | Published | Methods | Strength | |
|---|---|---|---|---|
Classical algorithm | SAN [5] | CVPR (2019) | Attention mechanism | High reconstruction quality |
DRN [10] | CVPR (2020) | Dual regression network | ||
DeFiAN [9] | IEEE Trans. Image Process. (2021) | Improved Hessian filter | ||
SwinIR [13] | CVPR (2021) | Transformer architecture | ||
DAT [7] | ICCV (2023) | Improved Transformer | ||
Lightweight algorithm | PAN [36] | ECCVW (2020) | Pixel attention | Low model complexity |
RFDN [18] | ECCV (2020) | Feature distillation connection | ||
MRDN [41] | Expert Syst. Appl. (2022) | Residual distillation | ||
MLRN [44] | CVPR (2023) | Mixer local residual | ||
VLESR [7] | Expert Syst. Appl. (2023) | Model pruning |
