A naive method to discover directions in the StyleGAN2 latent space

TL;DR

This paper presents a simple method to interpret and manipulate the latent space of StyleGAN2 for controlling facial features in generated images, with potential forensic applications.

Contribution

It introduces a naive approach to discover interpretable directions in StyleGAN2's latent space, linking them to biological facial features.

Findings

Successfully identified latent directions corresponding to facial features.

Demonstrated robustness of feature measures through statistical analysis.

Correlated latent directions with input perturbations.

Abstract

Several research groups have shown that Generative Adversarial Networks (GANs) can generate photo-realistic images in recent years. Using the GANs, a map is created between a latent code and a photo-realistic image. This process can also be reversed: given a photo as input, it is possible to obtain the corresponding latent code. In this paper, we will show how the inversion process can be easily exploited to interpret the latent space and control the output of StyleGAN2, a GAN architecture capable of generating photo-realistic faces. From a biological perspective, facial features such as nose size depend on important genetic factors, and we explore the latent spaces that correspond to such biological features, including masculinity and eye colour. We show the results obtained by applying the proposed method to a set of photos extracted from the CelebA-HQ database. We quantify some of these measures by utilizing two landmarking protocols, and evaluate their robustness through statistical analysis. Finally we correlate these measures with the input parameters used to perturb the latent spaces along those interpretable directions. Our results contribute towards building the groundwork of using such GAN architecture in forensics to generate photo-realistic faces that satisfy certain biological attributes.