This repository contains implementations of vanilla GAN, Least Squares GAN, Wasserstein GAN with Gradient Penalty, Spectral Normalization GAN and cGAN with projection discriminator. Everything was done as a part of this project.
We validated all models on a dataset of 25 gaussians (samples from WGAN-GP):
LIN dataset contains photographs of 41 proteins in fission yeast cells. To visualize similar proteins we used FID metric.
(left - real photographs, right - generated).
Quntatitative comparison of different models.
Samples from multichannel GAN (all details can be found here)
