Escaping from Collapsing Modes in a Constrained Space

TL;DR

This paper introduces BEGAN-CS, a modified GAN model with a latent-space constraint that enhances training stability, reduces mode collapse, and enables flexible image generation even with small datasets.

Contribution

We propose BEGAN-CS, a novel GAN variant that incorporates a latent-space constraint to improve stability and diversity without increasing complexity.

Findings

BEGAN-CS significantly reduces mode collapse.

The model maintains high image quality.

It performs well on small datasets and allows attribute-based image variations.

Abstract

Generative adversarial networks (GANs) often suffer from unpredictable mode-collapsing during training. We study the issue of mode collapse of Boundary Equilibrium Generative Adversarial Network (BEGAN), which is one of the state-of-the-art generative models. Despite its potential of generating high-quality images, we find that BEGAN tends to collapse at some modes after a period of training. We propose a new model, called \emph{BEGAN with a Constrained Space} (BEGAN-CS), which includes a latent-space constraint in the loss function. We show that BEGAN-CS can significantly improve training stability and suppress mode collapse without either increasing the model complexity or degrading the image quality. Further, we visualize the distribution of latent vectors to elucidate the effect of latent-space constraint. The experimental results show that our method has additional advantages of being able to train on small datasets and to generate images similar to a given real image yet with variations of designated attributes on-the-fly.