Skip to content

Latest commit

 

History

History

swinir

Folders and files

NameName
Last commit message
Last commit date

parent directory

..

LICENSE

LICENSE

 
 

README.md

README.md

 
 

input_classical.png

input_classical.png

 
 

input_color.png

input_color.png

 
 

input_gray.png

input_gray.png

 
 

input_jpeg.jpeg

input_jpeg.jpeg

 
 

input_lightweight.png

input_lightweight.png

 
 

input_real.png

input_real.png

 
 

output_classical.png

output_classical.png

 
 

swinir.py

swinir.py

 
 

README.md

SwinIR: Image Restoration Using Swin Transformer

Input

In case of classical model.
input

(Image from https://github.com/JingyunLiang/SwinIR/tree/main/testsets)

Output

In case of classical model.
Output

Usage

Automatically downloads the onnx and prototxt files on the first run. It is necessary to be connected to the Internet while downloading.

For the sample image,

$ python3 swinir.py --onnx

Please be careful that onnxruntime is used bacause ailia model is not implemented.

If you want to specify the input image, put the image path after the --input option.
You can use --savepath option to change the name of the output file to save.

$ python3 swinir.py --input IMAGE_PATH --savepath SAVE_IMAGE_PATH --onnx
(ex) $ python3 swinir.py --input input_classical.png --savepath example.png --onnx

By adding the --model_name option, you can choose the model.

$ python3 swinir.py --model_name MODEL_NAME --onnx
(ex) $ python3 swinir.py --model_name classical --onnx
(ex) $ python3 swinir.py --model_name lightweight --onnx
(ex) $ python3 swinir.py --model_name real --onnx
(ex) $ python3 swinir.py --model_name gray --onnx
(ex) $ python3 swinir.py --model_name color --onnx
(ex) $ python3 swinir.py --model_name jpeg --onnx

By adding the --video option, you can input the video. If you pass 0 as an argument to VIDEO_PATH, you can use the webcam input instead of the video file.

$ python3 swinir.py --video VIDEO_PATH --onnx
(ex) $ python3 swinir.py --video demo.mp4 --onnx
(ex) $ python3 swinir.py --video demo.mp4 -s output2.mp4 --onnx
(ex) $ python3 swinir.py --video demo.mp4 --model_name classical --onnx
(ex) $ python3 swinir.py --video demo.mp4 -s output.mp4 --model_name lightweight --onnx

Reference

Framework

Pytorch 1.7.1

Model Format

ONNX opset=11

Netron

001_classicalSR_DIV2K_s48w8_SwinIR-M_x2.onnx.prototxt 002_lightweightSR_DIV2K_s64w8_SwinIR-S_x2.onnx.prototxt 003_realSR_BSRGAN_DFO_s64w8_SwinIR-M_x4_GAN.onnx.prototxt 004_grayDN_DFWB_s128w8_SwinIR-M_noise25.onnx.prototxt 005_colorDN_DFWB_s128w8_SwinIR-M_noise25.onnx.prototxt 006_CAR_DFWB_s126w7_SwinIR-M_jpeg10.onnx.prototxt