Face Identity-Aware Disentanglement in StyleGAN

TL;DR

This paper introduces PluGeN4Faces, a plugin for StyleGAN that disentangles face attributes from identity, enabling attribute manipulation without affecting other facial characteristics.

Contribution

The paper presents a novel contrastive learning approach using movie frame images to improve face attribute disentanglement in StyleGAN.

Findings

Significantly reduces unintended identity changes during attribute editing.

Achieves better attribute manipulation fidelity compared to existing models.

Demonstrates effectiveness on diverse face images from movie frames.

Abstract

Conditional GANs are frequently used for manipulating the attributes of face images, such as expression, hairstyle, pose, or age. Even though the state-of-the-art models successfully modify the requested attributes, they simultaneously modify other important characteristics of the image, such as a person's identity. In this paper, we focus on solving this problem by introducing PluGeN4Faces, a plugin to StyleGAN, which explicitly disentangles face attributes from a person's identity. Our key idea is to perform training on images retrieved from movie frames, where a given person appears in various poses and with different attributes. By applying a type of contrastive loss, we encourage the model to group images of the same person in similar regions of latent space. Our experiments demonstrate that the modifications of face attributes performed by PluGeN4Faces are significantly less invasive on the remaining characteristics of the image than in the existing state-of-the-art models.