pokeGAN

Creats new kinds of pokemons using WGAN! (DCGAN is also supported)

dependence

cv2
tensorflow( >=1.0)
scipy
numpy

how to use

git clone ‘https://github.com/moxiegushi/pokeGAN.git’
cd pokeGAN
python resize.py
python RGBA2RGB.py
python pokeGAN.py

example pokemons

It is difficult to train a GAN perfectly, and as you can see someimages are meaningless. I'm very new to this. Please let me know if there are bugs :)