PyTorch implementation of VIGAN
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
datasets
imgs
models
options
util
README.md
test.py
train.py

README.md

VIGAN

This is a PyTorch implementation of the paper "VIGAN: Missing View Imputation with Generative Adversarial Networks". Please cite the paper in your publications if you find the source code useful to your research.

Installation

Install pytorch and torchvision.

VIGAN model

VIGAN is the model for imputing missing views based on generative adversarial networks which combines cross-domain relations given unpaired data with multi-view relations given paired data.

Dataset

Benchmark dataset: MNIST (http://deeplearning.net/data/mnist/mnist.pkl.gz).

Train the network to learn to generate digit images and the corresponding edges images of the digits images, inspired by CoGAN. We provide a python code to automaticlly download data and generate training data in the form of pairs of ( digit image, edges image ).

Train the model

When you train the model, you can tune the parameters in "options" folder.

python train.py

Test the model

python test.py

Acknowledgments

Code is inspired by CycleGAN and CoGAN.