VAE-GAN-in-tensorflow2

An implementation of VAE-GAN in tensorflow2

original paper:

http://arxiv.org/abs/1512.09300

input :

64×64×1 Grayscale image in tfrecord format.

batch size = 128;

I added a new idea in this code which did not appear in the original paper,it is called "weighted loss",which performed well in my task.

Get more information in my blog:

https://blog.csdn.net/qq_44065334/article/details/117982502