LoGANv2: Conditional Style-Based Logo Generation with Generative Adversarial Networks

TL;DR

LoGANv2 introduces a conditional style-based GAN architecture for logo generation, enhancing control and diversity of outputs while maintaining high quality, addressing challenges of multi-modality and interpretability in logo synthesis.

Contribution

The paper proposes a novel conditional extension to StyleGAN for logo synthesis, improving controllability and detail embedding in generated logos.

Findings

Conditional model produces more diverse outputs.

High-quality conditions enable finer detail control.

Unconditional model closely matches training distribution.

Abstract

Domains such as logo synthesis, in which the data has a high degree of multi-modality, still pose a challenge for generative adversarial networks (GANs). Recent research shows that progressive training (ProGAN) and mapping network extensions (StyleGAN) enable both increased training stability for higher dimensional problems and better feature separation within the embedded latent space. However, these architectures leave limited control over shaping the output of the network, which is an undesirable trait in the case of logo synthesis. This paper explores a conditional extension to the StyleGAN architecture with the aim of firstly, improving on the low resolution results of previous research and, secondly, increasing the controllability of the output through the use of synthetic class-conditions. Furthermore, methods of extracting such class conditions are explored with a focus on the human interpretability, where the challenge lies in the fact that, by nature, visual logo characteristics are hard to define. The introduced conditional style-based generator architecture is trained on the extracted class-conditions in two experiments and studied relative to the performance of an unconditional model. Results show that, whilst the unconditional model more closely matches the training distribution, high quality conditions enabled the embedding of finer details onto the latent space, leading to more diverse output.