NVlabs/noise2noise代码（一）概览与运行

我会带着你远行 2022-04-23 01:16 471阅读 0赞

**背景：**github上的noise2noise的官方代码是NVlabs的代码，我们希望在此基础上进行修改，所以需要初步看懂与运行。

**目的：**运行与跑通noise2noise的代码，训练与测试。

**论文地址**：[https://arxiv.org/abs/1803.04189][https_arxiv.org_abs_1803.04189]

**源码地址：**[https://github.com/NVlabs/noise2noise][https_github.com_NVlabs_noise2noise]

**论文详解**：[https://blog.csdn.net/weixin\_36474809/article/details/86535639][https_blog.csdn.net_weixin_36474809_article_details_86535639]

**目录**

一、环境要求与配置

1.1 用Anaconda配置python3.6 tf pip

1.2 安装依赖项

二、准备训练集与验证集

2.1 将ImageNet生成为训练集

2.2 用BSD300生成训练文件

2.3 Kodak验证集

三、训练与验证

3.1 训练

3.2 验证

3.3 预训练的网络

四、MRI图像去噪

Preparing the MRI training dataset

Training

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

# 一、环境要求与配置 #

作者github上源码需要安装python 3.6版本，运用Anaconda 5.2来配置相应的Python环境。创建相应的环境与配置。指令汇总如下，注意我们用的anaconda3的激活环境为 source acitvate n2n，去激活用source deactivate

conda create -n n2n python=3.6
    conda activate n2n
    conda install tensorflow-gpu
    python -m pip install --upgrade pip
    pip install -r requirements.txt

## 1.1 用Anaconda配置python3.6 tf pip ##

我们安装2018.12的anaconda版本，然后实现相应的配置。

**Ubuntu14.04安装Anaconda3-2018.12-x86\_64** [https://blog.csdn.net/weixin\_36474809/article/details/87804903][https_blog.csdn.net_weixin_36474809_article_details_87804903]

**运用Anaconda对python 3.6与tensorflow-gpu与pip环境配置** [https://blog.csdn.net/weixin\_36474809/article/details/87714182][https_blog.csdn.net_weixin_36474809_article_details_87714182]

按照上两篇博文配置好环境

## 1.2 安装依赖项 ##

pip install -r reauirements.txt

(n2n) jcx@smart-dsp:~/Desktop/xxr2019/NVlabs_noise2noise$ pip install -r requirements.txt

安装完成后，会出现下面信息

1.1.0->-r requirements.txt (line 4)) (1.15.4)
    Building wheels for collected packages: nibabel
      Building wheel for nibabel (setup.py) ... done
      Stored in directory: /home/jcx/.cache/pip/wheels/7e/97/9b/7cf05225afac39de2dacef758fc06c0c0bdf3e6efb40fabb9c
    Successfully built nibabel
    Installing collected packages: Pillow, typeguard, nibabel, scipy
      Found existing installation: scipy 1.2.0
        Uninstalling scipy-1.2.0:
          Successfully uninstalled scipy-1.2.0
    Successfully installed Pillow-5.2.0 nibabel-2.3.0 scipy-1.1.0 typeguard-2.2.2

# 二、准备训练集与验证集 #

程序中，**dataset\_tool\_tf.py**用于将相应的文件生成可用于训练n2n网络的tfrecords文件。

## 2.1 将ImageNet生成为训练集 ##

# This should run through roughly 50K images and output a file called `datasets/imagenet_val_raw.tfrecords`.
    python dataset_tool_tf.py 
    --input-dir "<path_to_imagenet>/ILSVRC2012_img_val" 
    --out=datasets/imagenet_val_raw.tfrecords

命令行中，--input-dir为输入的文件，--out为输出的文件

如果正确运行，类似于下面这样的输出：

<...long omitted...>
    49997 ./ImageNet/ILSVRC2012_img_val/ILSVRC2012_val_00002873.JPEG
    49998 ./ImageNet/ILSVRC2012_img_val/ILSVRC2012_val_00031550.JPEG
    49999 ./ImageNet/ILSVRC2012_img_val/ILSVRC2012_val_00009765.JPEG
    Dataset statistics:
      Formats:
        RGB: 49100 images
        L: 899 images
        CMYK: 1 images
      width,height buckets:
        >= 256x256: 48627 images
        < 256x256: 1373 images

## 2.2 用BSD300生成训练文件 ##

BSD300训练出的网络性能比ImageNet稍差，下载[BSD300 dataset][]

然后解压到相应文件夹。然后进行转换：

python dataset_tool_tf.py 
    --input-dir datasets/BSDS300-images/BSDS300/images/train 
    --out=datasets/bsd300.tfrecords

## 2.3 Kodak验证集 ##

Training tests validation loss against the [Kodak Lossless True Color Image Suite][] dataset

# Download the kodak validation set from http://r0k.us/graphics/kodak/
    python download_kodak.py --output-dir=datasets/kodak

# 三、训练与验证 #

## 3.1 训练 ##

在ImageNet上运用N2N的autoencoder

# try python config.py train --help for available options
    python config.py 
    --desc='-test' train 
    --train-tfrecords=datasets/imagenet_val_raw.tfrecords 
    --long-train=true 
    --noise=gaussian

然后可以用tensorboard查看相应的训练过程

cd results
    tensorboard --logdir .

训练过程相当耗时，例如高斯噪声的去噪网络，作者用ImageNet训练，在NVIDIA Titan V GPU上花费了将近7.5小时

**训练结束标志**：results/\* 文件夹之下生成一个network\_final.pickle的文件。

## 3.2 验证 ##

假定运行后结果存在 `results/00001-autoencoder-1gpu-L-n2n 之中`

命令行这样输入：

python config.py validate 
    --dataset-dir=datasets/kodak 
    --network-snapshot=results/00001-autoencoder-1gpu-L-n2n/network_final.pickle

## 3.3 预训练的网络 ##

泊松噪声和高斯噪声的预训练的网络如下：[https://drive.google.com/drive/folders/1-84ORv4wB8W3M6WngFTtccuW7SlPku0V][https_drive.google.com_drive_folders_1-84ORv4wB8W3M6WngFTtccuW7SlPku0V]

论文实验的结果：

训练

<table> 
 <thead> 
  <tr> 
   <th>Noise</th> 
   <th>Noise2Noise</th> 
   <th>Command line</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>Gaussian</td> 
   <td>Yes</td> 
   <td>python config.py train --noise=gaussian --noise2noise=true --long-train=true --train-tfrecords=datasets/imagenet_val_raw.tfrecords</td> 
  </tr> 
  <tr> 
   <td>Gaussian</td> 
   <td>No</td> 
   <td>python config.py train --noise=gaussian --noise2noise=false --long-train=true --train-tfrecords=datasets/imagenet_val_raw.tfrecords</td> 
  </tr> 
  <tr> 
   <td>Poisson</td> 
   <td>Yes</td> 
   <td>python config.py train --noise=poisson --noise2noise=true --long-train=true --train-tfrecords=datasets/imagenet_val_raw.tfrecords</td> 
  </tr> 
  <tr> 
   <td>Poisson</td> 
   <td>No</td> 
   <td>python config.py train --noise=poisson --noise2noise=false --long-train=true --train-tfrecords=datasets/imagenet_val_raw.tfrecords</td> 
  </tr> 
 </tbody> 
</table>

验证：

<table> 
 <thead> 
  <tr> 
   <th>Noise</th> 
   <th>Dataset</th> 
   <th>Command line</th> 
   <th>Expected PSNR (dB)</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>Gaussian</td> 
   <td>kodak</td> 
   <td>python config.py validate --dataset-dir=datasets/kodak --noise=gaussian --network-snapshot=&lt;.../network_final.pickle&gt;</td> 
   <td>32.38 (n2c) / 32.39 (n2n)</td> 
  </tr> 
  <tr> 
   <td>Gaussian</td> 
   <td>bsd300</td> 
   <td>python config.py validate --dataset-dir=datasets/bsd300 --noise=gaussian --network-snapshot=&lt;.../network_final.pickle&gt;</td> 
   <td>31.01 (n2c) / 31.02 (n2n)</td> 
  </tr> 
  <tr> 
   <td>Poisson</td> 
   <td>kodak</td> 
   <td>python config.py validate --dataset-dir=datasets/kodak --noise=poisson --network-snapshot=&lt;.../network_final.pickle&gt;</td> 
   <td>31.66 (n2c) / 31.66 (n2n)</td> 
  </tr> 
  <tr> 
   <td>Poisson</td> 
   <td>bsd300</td> 
   <td>python config.py validate --dataset-dir=datasets/bsd300 --noise=poisson --network-snapshot=&lt;.../network_final.pickle&gt;</td> 
   <td>30.27 (n2c) / 30.26 (n2n)</td> 
  </tr> 
 </tbody> 
</table>

# 四、MRI图像去噪 #

### Preparing the MRI training dataset ###

Use the `dataset_tool_mri.py` script to generate training and validation datasets for the N2N MRI case.

**Step \#1**: Download the IXI-T1 dataset from: [https://brain-development.org/ixi-dataset/][https_brain-development.org_ixi-dataset]. Unpack to some location.

**Step \#2**: Convert the IXI-T1 dataset into a set of PNG files:

# Assumes you have downloaded and untarred IXI-T1 under ~/Downloads/IXI-T1.
    
    python dataset_tool_mri.py genpng --ixi-dir=~/Downloads/IXI-T1 --outdir=datasets/ixi-png

**Step \#3**: Convert a subset of the IXI-T1 dataset into training and validation sets:

python dataset_tool_mri.py genpkl --png-dir=datasets/ixi-png --pkl-dir=datasets

### Training ###

python config_mri.py

A successful invocation should output the following:

dnnlib: Running train_mri.train() on localhost...
    Loading training set.
    Loading dataset from datasets\ixi_train.pkl
    <...long log omitted...>
    Epoch 297/300: time=107.981, train_loss=0.0126064, test_db_clamped=31.72174, lr=0.000002
    Epoch 298/300: time=107.477, train_loss=0.0125972, test_db_clamped=31.73622, lr=0.000001
    Epoch 299/300: time=106.927, train_loss=0.0126012, test_db_clamped=31.74232, lr=0.000001
    Saving final network weights.
    Resetting random seed and saving a bunch of example images.
    dnnlib: Finished train_mri.train() in 8h 59m 19s.

The expected average PSNR on the validation set (named `test_db_clamped` in code) is roughly 31.74 dB.

Noise-to-noise training is enabled by default for the MRI case. To use noise-to-clean training, edit `config_mri.py` and change `corrupt_targets=True` to `corrupt_targets=False`.

Training for 300 epochs takes roughly 9 hours on an NVIDIA Titan V GPU.

[https_arxiv.org_abs_1803.04189]: https://arxiv.org/abs/1803.04189
[https_github.com_NVlabs_noise2noise]: https://github.com/NVlabs/noise2noise
[https_blog.csdn.net_weixin_36474809_article_details_86535639]: https://blog.csdn.net/weixin_36474809/article/details/86535639
[https_blog.csdn.net_weixin_36474809_article_details_87804903]: https://blog.csdn.net/weixin_36474809/article/details/87804903
[https_blog.csdn.net_weixin_36474809_article_details_87714182]: https://blog.csdn.net/weixin_36474809/article/details/87714182
[BSD300 dataset]: https://www2.eecs.berkeley.edu/Research/Projects/CS/vision/bsds/BSDS300-images.tgz
[Kodak Lossless True Color Image Suite]: http://r0k.us/graphics/kodak/
[https_drive.google.com_drive_folders_1-84ORv4wB8W3M6WngFTtccuW7SlPku0V]: https://drive.google.com/drive/folders/1-84ORv4wB8W3M6WngFTtccuW7SlPku0V
[https_brain-development.org_ixi-dataset]: https://brain-development.org/ixi-dataset/