Mutations in GAN

By Pawel Kopciewicz & Vitalii Morskyi

The evolutionary impact on the Generative Adversarial Network was studied in this repo. The traditional gradient descent was combined with mutations and selection to improve the process of local minima escape.

This is the repository for the following article: Exploration of the evolution-based footprint on the Generative Adversarial Networks

Acknowledgments

This code borrows heavily from the following sources:

• Erik Linder-Norén PyTorch-GAN .
• Erik Linder-Norén Keras-GAN .
• TensorFlow DCGAN

The idea of the project was inspired by this paper: Evolutionary Generative Adversarial Networks (E-GAN) .