galerkin / AGE

Code for paper "Adversarial Generator-Encoder Networks" by Dmitry Ulyanov, Andrea Vedaldi and Victor Lempitsky

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.

Pretrained models

1. First install dev version of pyTorch (see manual here) and make sure you have jupyter notebook ready.

2. Then download the models with the script:

bash download_pretrained.sh

3. Run jupyter notebook and go through evaluate.ipynb.

Here is an example of samples and reconstructions for imagenet, celeba and cifar10 datasets.

Celeba

Samples	Reconstructions

Cifar10

Samples	Reconstructions

Tiny ImageNet

Samples	Reconstructions