Feature Proliferation -- the "Cancer" in StyleGAN and its Treatments

TL;DR

This paper identifies a phenomenon called Feature Proliferation in StyleGAN that causes artifacts and reduces image diversity, and proposes a feature rescaling method to mitigate these issues more effectively than the truncation trick.

Contribution

The paper discovers Feature Proliferation in StyleGAN and introduces a novel feature rescaling technique to improve image quality and diversity.

Findings

Feature Proliferation causes artifacts in StyleGAN images.

The proposed method better preserves image features than truncation.

Experimental results validate the effectiveness of the new approach.

Abstract

Despite the success of StyleGAN in image synthesis, the images it synthesizes are not always perfect and the well-known truncation trick has become a standard post-processing technique for StyleGAN to synthesize high-quality images. Although effective, it has long been noted that the truncation trick tends to reduce the diversity of synthesized images and unnecessarily sacrifices many distinct image features. To address this issue, in this paper, we first delve into the StyleGAN image synthesis mechanism and discover an important phenomenon, namely Feature Proliferation, which demonstrates how specific features reproduce with forward propagation. Then, we show how the occurrence of Feature Proliferation results in StyleGAN image artifacts. As an analogy, we refer to it as the" cancer" in StyleGAN from its proliferating and malignant nature. Finally, we propose a novel feature rescaling method that identifies and modulates risky features to mitigate feature proliferation. Thanks to our discovery of Feature Proliferation, the proposed feature rescaling method is less destructive and retains more useful image features than the truncation trick, as it is more fine-grained and works in a lower-level feature space rather than a high-level latent space. Experimental results justify the validity of our claims and the effectiveness of the proposed feature rescaling method. Our code is available at https://github. com/songc42/Feature-proliferation.