SZU-AdvTech-2023/200-Zero-Shot-Noise2Noise-Efficient-Image-Denoising-without-Any-Data

Resources

• Arxiv
• Conference
• Supplementary

Python Requirements

This code was tested on:

• Python 3.7
• Pytorch 1.7

Training without using ground-truth data

run

python main.py -i {noisy_image_path}

The clean image will be generated under /output.