By Zhi-Song, Li-Wen Wang, Chu-Tak Li, Marie-Paule Cani and Wan-Chi Siu
This repo only provides simple testing codes, pretrained models and the network strategy demo.
We propose a joint image denoising and Super-Resolution model by using generative Variational AutoEncoder (dSRVAE)
We participate CVPRW NTIRE2020 Real Image Super-Resolution Challenge
Please check our paper
@InProceedings{Liu2020dsrvae,
author = {Liu, Zhi-Song Siu, Wan-Chi and and Wang, Li-Wen and Li, Chu-Tak and Marie-Paule Cani and Yui-Lam Chan},
title = {Unsupervised Real Image Super-Resolution via Generative Variational AutoEncoder},
booktitle = {IEEE International Conference on Computer Vision and Pattern Recognition Workshop(CVPRW)},
month = {June},
year = {2020}
}
• First working on using Variational AutoEncoder for image denoising.
• Then the Super-Resolution Sub-Network (SRSN) is attached as a small overhead to the DAE which forms the proposed dSRVAE to output super-resolved images.
Python > 3.0
OpenCV library
Pytorch > 1.0
NVIDIA GPU + CUDA
The complete architecture is shown as follows,
- Download pre-trained models from
https://drive.google.com/open?id=1SUZGE04vw5_yDYiw6PJ4sbHAOIEV6TJ7
and copy them to the folder "models"
- Copy your image to folder "Test" and run
$ python test.py
The SR images will be in folder "Result" 3. For self-ensemble, run
$ python test_enhance.py
s1. Testing images on NTIRE2020 Real World Super-Resolution Challenge - Track 1: Image Processing artifacts can be downloaded from the following link:
https://drive.google.com/open?id=10ZutE-0idGFW0KUyfZ5-2aVSiA-1qUCV
s2. Testing images on NTIRE2020 Real World Super-Resolution Challenge - Track 2: Smartphone Images can be downloaded from the following link:
https://drive.google.com/open?id=1_R4kRO_029g-HNAzPobo4-xwp86bMZLW
s3. Validation images on NTIRE2020 Real World Super-Resolution Challenge - Track 1 and Track 2 can be downloaded from the following link:
https://drive.google.com/open?id=1nKEJ4N2V-0NFicfJxm8AJqsjXoGMYjMp
https://competitions.codalab.org/competitions/22220#learn_the_details
- Train the Denoising VAE by running
$ python main_denoiser.py
- Train the super-resolution SRSN overhead by running
$ python main_GAN.py
Results on 4x image SR on Track 1 dataset
Special thanks to the contributions of Jakub M. Tomczak for their VAE with a VampPrior on KL loss calculation.