You have two before/after images with the same size.
Place the restored image and noised image into specific folder, respectively.
Before running the gif_maker.py, you can adjust the following parameters:
--frame_dir : the folder of being generated essential frames
--resize_scale: if image is too larger, you can consider to resize
--step : each frame with [step] pixels
--type: : png or jpg or others
--frequency : spending seconds in single loop
--repeat_GIF : 0: repeat, 1: no repeat
--result_dir : path of result gif
Restoration (visual result)
| Restoration tasks | Restored images |
|---|---|
| Denoise (real-world) |  |
| Derain (rainstreak) |  |
| Derain (raindrop) |  |
| Dehaze (densehaze) |  |
| Deblur (motionblur) |  |
| LLEnhancement |  |
| Retouching |  |
Inpainting (visual result)
| Restoration tasks | Restored images |
|---|---|
| Inpainting |  |
Segmentation (visual result)
| Restoration tasks | Restored images |
|---|---|
| Instance segmentation |  |
| Semantic segmentation |  |
| K-mean |  |