Github author: 蔡詠平
1.face detect and cut the ROI in the image (Data/face_detect.py)
2.random flip the image left to right
3.train the model by BEGAN algo (began5.py)
4.generate image (generate2.py)
I try a lot of paramater this one is the best one in 48000 iter.
Boundary Equilibrium Generative Adversarial Networks: https://arxiv.org/abs/1703.10717
DCGAN: https://github.com/chenyuntc/pytorch-book/tree/master/chapter07-AnimeGAN
BEGAN implement: https://github.com/sunshineatnoon/Paper-Implementations/tree/master/BEGAN