This paper presents a new depth network image denoising method .  It is different from the existing denoising methods based on depth network , We will have a new ACP Item is introduced into the optimization problem , Then use the optimization process , Design a deep network by expanding strategies .  our ACP The driving denoising network combines some valuable results of classical denoising methods , And to a certain extent, it improves its interpretability .

Official code ：

1 Introduce

### Contents

Dataset| Contains three folders (train, test and Benchmark_test), You can use train Data sets placed in train in , Put the test data set in test in , take SIDD/DnD Benchmark on Benchmark_test in
Deam_models| When you want to test | when , Pre trained models
real| Some about real image denoising python file
statistics| Record the results during training

### Training on AWGN
We need to retrain our network , Please put your own training data set in './Dataset/train ', And then run ' train.py’
### Training on real-world noise
Retrain our network to remove real-world noise : Download the training data set to `./Dataset/train` And use './Dataset/train/gen_dataset_real.py' Pack them into h5py Format .
You can start your https://www.eecs.yorku.ca/~kamel/sidd/dataset.php

and http://ani.stat.fsu.edu/~abarbu/Renoir.html Get data set
Set the training and testing path to your own path , And run ' train.py '. More details , Please refer to https://github.com/JimmyChame/SADNet
### Testing on AWGN
Test your own image , Put your data set in ' Dataset/test/your_test_name ' in , And then run ' Synthetic_test.py '
### Testing on real-world noise
To test real-world noise data sets : Download the test data set to './Dataset/Benchmark_test ' And run
'Benchmark_test.py '
You can learn from https://www.eecs.yorku.ca/~kamel/sidd/benchmark.php and https://noise.visinf.tu-darmstadt.de/benchmark/ Acquired data set .

----------------------------------------------------------------------------------

First run train.py

D:\ProgramData\Anaconda3\envs\python36\python.exe "D:/Papers to read/2022.07/Adaptive Consistency Prior based Deep Network for Image_Denoising/DeamNet-main/DeamNet-main/train.py"
Traceback (most recent call last):
  File "D:/Papers to read/2022.07/Adaptive Consistency Prior based Deep Network for Image_Denoising/DeamNet-main/DeamNet-main/train.py", line 15, in <module>
    from real_dataloader import *
  File "D:\Papers to read\2022.07\Adaptive Consistency Prior based Deep Network for Image_Denoising\DeamNet-main\DeamNet-main\real_dataloader.py", line 6, in <module>
    import h5py
ModuleNotFoundError: No module named 'h5py'

Process finished with exit code 1

Error reason ： install h5py library

Download Database , Chose the smallest one （ Both 1.6GB）. If the operation is successful , Next, prepare to learn ‘Dataset.py’, Combine the code to learn the content of the article .