| Class | Task | Train/Test | Description |
|---|---|---|---|
| PairedImageDataset | Image Super-Resolution | Train | Support paired data |
| SingleImageDataset | Image Super-Resolution | Test | Only read low quality images, used in tests without Ground-Truth |
| REDSDataset | Video Super-Resolution | Train | REDS training dataset |
| Vimeo90KDataset | Video Super-Resolution | Train | Vimeo90K training dataset |
| VideoTestDataset | Video Super-Resolution | Test | Base video test dataset, supporting Vid4, REDS testing datasets |
| VideoTestVimeo90KDataset | Video Super-Resolution | Test | Inherit VideoTestDataset, Vimeo90K testing dataset |
| VideoTestDUFDataset | Video Super-Resolution | Test | Inherit VideoTestDataset, testing dataset for method DUF, supporting Vid4 dataset |
| FFHQDataset | Face Generation | Train | FFHQ training dataset |
- Common transformations and functions are in transforms.py and util.py, respectively