LC-GAN

Requirements

For datasets, you can download the CelebA dataset here

Example scripts are provided for easy use.

File	Usage
`run.sh`	Train a VAE
`extract.sh`	Save the posterior of the images in the datset using a trained VAE in `run.sh`
`real.sh`	Train a realism classifier without using GAN
`real_gan.sh`	Train a realism classifier using GAN
`attr.sh`	Train a attribute classifier without using GAN
`generate_real.sh`	Generate realistic images using VAE, realism classifier or realism actor
`transform.sh`	Perform identity-preserved transformation using VAE, realism classifier and attribute classifier

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
models		models
utils		utils
.gitignore		.gitignore
README.md		README.md
attr.sh		attr.sh
extract.sh		extract.sh
extract_latent.py		extract_latent.py
generate_real.sh		generate_real.sh
real.sh		real.sh
real_gan.sh		real_gan.sh
run.sh		run.sh
sample_real.py		sample_real.py
test.py		test.py
train_attr.py		train_attr.py
train_real.py		train_real.py
train_real_gan.py		train_real_gan.py
train_vae.py		train_vae.py
transform.sh		transform.sh
transform_attr.py		transform_attr.py