Coevolution of Generative Adversarial Networks

TL;DR

This paper introduces COEGAN, a coevolutionary approach combining neuroevolution with GAN training to improve stability, automate architecture design, and address mode collapse, demonstrated on MNIST.

Contribution

It presents a novel coevolutionary framework for GANs that enhances training stability and automates architecture discovery, partially mitigating mode collapse.

Findings

Improved training stability of GANs.

Automatic discovery of efficient architectures.

Partial solution to mode collapse.

Abstract

Generative adversarial networks (GAN) became a hot topic, presenting impressive results in the field of computer vision. However, there are still open problems with the GAN model, such as the training stability and the hand-design of architectures. Neuroevolution is a technique that can be used to provide the automatic design of network architectures even in large search spaces as in deep neural networks. Therefore, this project proposes COEGAN, a model that combines neuroevolution and coevolution in the coordination of the GAN training algorithm. The proposal uses the adversarial characteristic between the generator and discriminator components to design an algorithm using coevolution techniques. Our proposal was evaluated in the MNIST dataset. The results suggest the improvement of the training stability and the automatic discovery of efficient network architectures for GANs. Our model also partially solves the mode collapse problem.