This repository contains code for the paper
"Adversarial Generator-Encoder Networks" by Dmitry Ulyanov, Andrea Vedaldi, Victor Lempitsky.
For now only evaluation code and the models from paper are here, training code will be added later this week.
-
First install dev version of pyTorch (see manual here) and make sure you have
jupyter notebook
ready. -
Then download the models with the script:
bash download_pretrained.sh
- Run
jupyter notebook
and go throughevaluate.ipynb
.
Here is an example of samples and reconstructions for imagenet
, celeba
and cifar10
datasets.
Samples | Reconstructions |
---|---|
![]() |
![]() |
Samples | Reconstructions |
---|---|
![]() |
![]() |
Samples | Reconstructions |
---|---|
![]() |
![]() |