LoGAN: Generating Logos with a Generative Adversarial Neural Network Conditioned on color

TL;DR

LoGAN is a novel conditional GAN model that generates logos based on color, demonstrating promising results in aiding designers with automated logo creation.

Contribution

This paper introduces LoGAN, an improved auxiliary classifier Wasserstein GAN that conditions logo generation on color, addressing the challenges of logo diversity and categorical properties.

Findings

Achieved 0.8 precision and 0.7 recall in color-conditioned logo generation.

Generated 768 logo instances across 12 color classes.

Demonstrated potential for AI-assisted logo design.

Abstract

Designing a logo is a long, complicated, and expensive process for any designer. However, recent advancements in generative algorithms provide models that could offer a possible solution. Logos are multi-modal, have very few categorical properties, and do not have a continuous latent space. Yet, conditional generative adversarial networks can be used to generate logos that could help designers in their creative process. We propose LoGAN: an improved auxiliary classifier Wasserstein generative adversarial neural network (with gradient penalty) that is able to generate logos conditioned on twelve different colors. In 768 generated instances (12 classes and 64 logos per class), when looking at the most prominent color, the conditional generation part of the model has an overall precision and recall of 0.8 and 0.7 respectively. LoGAN's results offer a first glance at how artificial intelligence can be used to assist designers in their creative process and open promising future directions, such as including more descriptive labels which will provide a more exhaustive and easy-to-use system.