Installation of EfficientDerain

This is the installation, training and validating guide for EfficientDerain network in the paper "EfficientDeRain: Learning Pixel-wise Dilation Filtering for High-Efficiency Single-Image Deraining " (https://arxiv.org/abs/2009.09238)

Installation

Run the script below to install (Conda):

conda create -n proposed1 python=3.6.7
conda activate proposed1
conda install -y pytorch=1.4.0 cudatoolkit=10.0 torchvision -c pytorch
conda install h5py opencv
pip install tensorboardX scikit-image==0.17.2 pytorch-msssim==0.2.1

Datasets

To train and evaluate the models, please download training and testing datasets from https://drive.google.com/file/d/1IBENJqN6XU5HMlSQ2W7jkdW22Z1x4W9K/view?usp=sharing and place the unzipped folders into the project folder.

Pretrained models

Here is the urls of pretrained models (includes v3_rain100H, v3_rain1400, v3_SPA, v4_rain100H, v4_rain1400, v4_SPA) :

direct download: http://www.xujuefei.com/models_effderain.zip

google drive: https://drive.google.com/file/d/1OBAIG4su6vIPEimTX7PNuQTxZDjtCUD8/view?usp=sharing

baiduyun: https://pan.baidu.com/s/1kFWP-b3tD8Ms7VCBj9f1kw (pwd: vr3g)

Train

• The code shown corresponds to version v3, for v4 change the value of argument "rainaug" in file "./train_*.sh" to the "true" (train_*.sh means it's the training script of dataset *)
• Unzip the "Streaks_Garg06.zip" in the "./rainmix"
• Change the value of argument "baseroot" in file "./train.sh" to the path of training data (Crop size must be divisible by 16)
• Edit the function "get_files" in file "./utils" according to the format of the training data
• Execute

sh train___.sh

Test

• The code shown corresponds to version v3
• Change the value of argument "load_name" in file "./test.sh" to the path of pretained model
• Change the value of argument "baseroot" in file "./test.sh" to the path of testing data
• Edit the function "get_files" in file "./utils" according to the format of the testing data
• Execute

sh test.sh

Bibtex

@inproceedings{guo2020efficientderain,
      title={EfficientDeRain: Learning Pixel-wise Dilation Filtering for High-Efficiency Single-Image Deraining}, 
      author={Qing Guo and Jingyang Sun and Felix Juefei-Xu and Lei Ma and Xiaofei Xie and Wei Feng and Yang Liu},
      year={2021},
      booktitle={AAAI}
}