ESRGAN-Keras

Recurring the ESRGAN(https://arxiv.org/abs/1809.00219) with Keras

It was programed based on https://github.com/MathiasGruber/SRGAN-Keras, and refered from https://github.com/SavaStevanovic/ESRGAN

It is not finished totally, there are several bugs, so please don't directly use ESRGAN.py.
You'd butter only regard it as a reference!!! Especially the weights of losses!!
I really don't know which number is able to use.
=_= -> QAQ -> orz

Environment: Python 3.6 + Keras 2.2.4 + Tensorflow 1.12 + PyCharm 2018

I have upload the weights of my generator model(RRDB). You can use it after copying my generator model code. If you don't copy the model code, it may report some errors beacause I used 'tf.xxx' in my model.

To be honest, I recommand you to train your own discriminator due to the discriminator has several types and visions, and I'm not sure which one is better. You can add a Dropout(0.4) layer at the last of the discriminator to keep the train process stable.

My recurrence doesn't use RaGAN due to bugs. I think my code maybe have some bugs I couldn't understand.

What's more, I use DIV2K datasets only. After doing experiments, I'm sure that 'the more high quality data, the better model performance' is TRUE.

the examples of my ESRGAN(without RaGAN)

Baboon in Set14

Zebra in Set14

The next two figures show how ESRGAN directly super-resolute the actual natural image. These two images are cropped from the original image of the DIV2K dataset. There is no "super-resolution original image" that can be compared, so their effects can truly reflect the actual application effect of super-resolution, rather than the reconstruction effect.

002-(4,5) in DIV2K

050-(2,2) in DIV2K

Other pictures' PSNR and SSIM are higher than these, but I think it is more clear. Don't mind it too much if you don't want to use it on security field and medical field.