This repository serves as the official test implementation of my paper, which has been preprinted on arXiv.
- Xie, X., Wu, Y., Ni, H. and He, C., 2023. NODE-ImgNet: a PDE-informed effective and robust model for image denoising. arXiv preprint arXiv:2305.11049.
The provided code in this repository corresponds to the concepts and methodologies described in the paper.
Requirements for a successful implementation of the codes can be found in requirements.txt
.
The training dataset of Gaussian noisy is downloaded at https://kedema.org/project/exploration/index.html
After downloading, place the folder in the ./data/GaussianTrainingData/ directory.
Below are some example commands for training or testing NODE-ImgNet:
python gaussian_gray_denoising.py --noise_level 25
python gaussian_color_denoising.py --noise_level 25
python gaussian_gray_denoising.py --is_blind
python gaussian_color_denoising.py --is_blind
If you want to cite this paper, please refer to the following format
-
Xie, X., Wu, Y., Ni, H. and He, C., 2023. NODE-ImgNet: a PDE-informed effective and robust model for image denoising. arXiv preprint arXiv:2305.11049.
-
@article{xie2023node,
title={NODE-ImgNet: a PDE-informed effective and robust model for image denoising},
author={Xie, Xinheng and Wu, Yue and Ni, Hao and He, Cuiyu},
journal={arXiv preprint arXiv:2305.11049},
year={2023} }