JavanTang / AECR-Net

Contrastive Learning for Compact Single Image Dehazing, CVPR2021

AECR-Net

Contrastive Learning for Compact Single Image Dehazing, CVPR2021. Official Pytorch based implementation.

Paper

arxiv

Pytorch Version

• model
• CR loss
• pretrained models

MindSpore Version

https://github.com/Booooooooooo/AECRNet-MindSpore by @wyb

Performance

Others

Pretrained models:

https://pan.baidu.com/s/13crsXwwhkI5A3MlHtPihuA password: xhyi

22-09-25 Update

Directory Tree

.
├── create_dataset.py
├── datasets
│   ├── ITS_test
│   └── ITS_train
├── data_utils
│   ├── DH.py
│   ├── ITS_h5.py
│   ├── NH.py
│   └── __pycache__
├── img
│   ├── aecrnet.png
│   ├── example.png
│   ├── performance.png
│   └── trade_off.png
├── ITS_v2
│   ├── clear
│   ├── clear.zip
│   ├── hazy
│   ├── hazy.zip
│   ├── trans
│   └── trans.zip
├── logs
│   ├── ITS_train_cdnet_test
│   ├── its_train_ffa_test
│   └── ITS_train_ffa_test
├── logs_train
│   ├── args_ITS_train_cdnet_test.txt
│   ├── args_its_train_ffa_test.txt
│   ├── args_ITS_train_ffa_test.txt
│   └── ITS_train_cdnet_test.txt
├── metrics.py
├── models
│   ├── AECRNet.py
│   ├── CR.py
│   ├── DCNv2
│   ├── DCNv2.zip
│   ├── deconv.py
│   └── __pycache__
├── numpy_files
├── option.py
├── README.md
├── requirements
├── samples
│   ├── ITS_train_cdnet_test
│   ├── its_train_ffa_test
│   └── ITS_train_ffa_test
├── test.py
├── train_aecrnet.py
└── trained_models
    ├── ITS_train_cdnet_test.pk
    └── ITS_train_cdnet_test.pk.best

Dataset

refer to: https://sites.google.com/view/reside-dehaze-datasets/reside-v0

ITS (Indoor Training Set)：
（Dropbox）： http://t.cn/RHjBQIV 
（Baidu Yun）：https://pan.baidu.com/s/16rm4zUF8uVRs3Ux5T9CMMA    Passward:  tqyh

put on ./ITS_v2, and run python create_dataset.py to create h5 files.