Towards Disentangling Latent Space for Unsupervised Semantic Face Editing

TL;DR

This paper introduces STGAN-WO, a novel unsupervised method for disentangling latent space in StyleGAN to enable independent semantic face editing without labeled data.

Contribution

The paper proposes STGAN-WO, combining weight decomposition, orthogonal regularization, and structure-texture independent architecture for improved unsupervised semantic face editing.

Findings

STGAN-WO achieves better attribute editing than state-of-the-art methods.

Unsupervised disentanglement of facial attributes is effective with the proposed architecture.

The method allows independent control of texture and structure in face synthesis.

Abstract

Facial attributes in StyleGAN generated images are entangled in the latent space which makes it very difficult to independently control a specific attribute without affecting the others. Supervised attribute editing requires annotated training data which is difficult to obtain and limits the editable attributes to those with labels. Therefore, unsupervised attribute editing in an disentangled latent space is key to performing neat and versatile semantic face editing. In this paper, we present a new technique termed Structure-Texture Independent Architecture with Weight Decomposition and Orthogonal Regularization (STIA-WO) to disentangle the latent space for unsupervised semantic face editing. By applying STIA-WO to GAN, we have developed a StyleGAN termed STGAN-WO which performs weight decomposition through utilizing the style vector to construct a fully controllable weight matrix to regulate image synthesis, and employs orthogonal regularization to ensure each entry of the style vector only controls one independent feature matrix. To further disentangle the facial attributes, STGAN-WO introduces a structure-texture independent architecture which utilizes two independently and identically distributed (i.i.d.) latent vectors to control the synthesis of the texture and structure components in a disentangled way. Unsupervised semantic editing is achieved by moving the latent code in the coarse layers along its orthogonal directions to change texture related attributes or changing the latent code in the fine layers to manipulate structure related ones. We present experimental results which show that our new STGAN-WO can achieve better attribute editing than state of the art methods.