Generative Adversarial Networks

An 70 lines Knet implementation of MLP GAN for MNIST data.

Usage

Install require packages:

Pkg.add("ImageMagick")
Pkg.add("Knet")

Currently, the latest version of Knet is required to run this, get it via:

Pkg.checkout("Knet")

Run the gan:

julia gan.jl

Unofficial implementation of StyleGan2 paper

Coloring black and white images with deep learning.

Py Library - Vanilla GAN, DCGAN, WGAN, SNGAN, SRGAN, StyleGAN, VAE, Pix2PixHD

A denoising autoencoder + adversarial losses and attention mechanisms for face swapping.