xinntao/ESRGAN
1
1
Fork 0
mirror of https://github.com/xinntao/ESRGAN.git synced 2024-08-17 08:30:32 +03:00
Code Issues Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Xintao 2018-09-06 14:19:15 +08:00 committed by GitHub
parent 917c10a487
commit 74c6cd1a63
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 11 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
README.md
View File

@ -1,4 +1,4 @@
# ESRGAN (Enhanced SRGAN) [[Paper]](https://arxiv.org/abs/1809.00219) [[BasicSR]](https://github.com/xinntao/BasicSR)
# ESRGAN (Enhanced SRGAN) [[Paper]](https://arxiv.org/abs/1809.00219) [[BasicSR]](https://github.com/xinntao/BasicSR)
### Enhanced Super-Resolution Generative Adversarial Networks
By Xintao Wang, [Ke Yu](https://yuke93.github.io/), Shixiang Wu, [Jinjin Gu](http://www.jasongt.com/), Yihao Liu, [Chao Dong](https://scholar.google.com.hk/citations?user=OSDCB0UAAAAJ&hl=en), [Chen Change Loy](http://personal.ie.cuhk.edu.hk/~ccloy/), [Yu Qiao](http://mmlab.siat.ac.cn/yuqiao/), [Xiaoou Tang](https://scholar.google.com/citations?user=qpBtpGsAAAAJ&hl=en)
@ -8,7 +8,7 @@ This repo only provides simple testing codes, pretrained models and the netwrok
We won the first place in [PIRM2018-SR competition](https://www.pirm2018.org/PIRM-SR.html) (region 3) and got the best perceptual index.
The paper is accepted to [ECCV2018 PIRM Workshop](https://pirm2018.org/).
### BibTeX
#### BibTeX
@article{wang2018esrgan,
author={Wang, Xintao and Yu, Ke and Wu, Shixiang and Gu, Jinjin and Liu, Yihao and Dong, Chao and Loy, Chen Change and Qiao, Yu and Tang, Xiaoou},
@ -18,7 +18,7 @@ The paper is accepted to [ECCV2018 PIRM Workshop](https://pirm2018.org/).
}
<p align="center">
<img height="400" src="figures/baboon.png">
<img height="400" src="figures/baboon.jpg">
</p>
The **RRDB_PSNR** PSNR_oriented model trained with DF2K dataset (a merged dataset with [DIV2K](https://data.vision.ee.ethz.ch/cvl/DIV2K/) and [Flickr2K](http://cv.snu.ac.kr/research/EDSR/Flickr2K.tar) (proposed in [EDSR](https://github.com/LimBee/NTIRE2017))) is also able to achive high PSNR performance.
@ -69,17 +69,17 @@ We improve the [SRGAN](https://arxiv.org/abs/1609.04802) from three aspects:
In contrast to SRGAN, which claimed that **deeper models are increasingly difficult to train**, our deeper ESRGAN model shows its superior performance with easy training.
<p align="center">
<img height="100" src="figures/architecture.png">
<img height="100" src="figures/architecture.jpg">
</p>
<p align="center">
<img height="130" src="figures/RRDB.png">
<img height="130" src="figures/RRDB.jpg">
</p>
## Network Interpolation
We propose the **network interpolation strategy** to balance the visual quality and PSNR.
<p align="center">
<img height="500" src="figures/net_interp.png">
<img height="500" src="figures/net_interp.jpg">
</p>
We show the smooth animation with the interpolation parameters changing from 0 to 1.
@ -94,20 +94,15 @@ We show the smooth animation with the interpolation parameters changing from 0 t
PSNR (evaluated on the luminance channel in YCbCr color space) and the perceptual index used in the PIRM-SR challenge are also provided for reference.
<p align="center">
<img src="figures/rlt_01.png">
<img src="figures/qualitative_cmp_01.jpg">
</p>
<p align="center">
<img src="figures/rlt_02.png">
<img src="figures/qualitative_cmp_02.jpg">
</p>
<p align="center">
<img src="figures/rlt_03.png">
<img src="figures/qualitative_cmp_03.jpg">
</p>
<p align="center">
<img src="figures/rlt_04.png">
</p>
<p align="center">
<img src="figures/rlt_05.png">
</p>
<p align="center">
<img src="figures/rlt_06.png">
<img src="figures/qualitative_cmp_04.jpg">
</p>