FaceController: Controllable Attribute Editing for Face in the Wild

TL;DR

FaceController introduces a simple, efficient feed-forward network for high-fidelity face attribute editing in the wild, enabling diverse manipulations like swapping, relighting, and makeup transfer with disentangled control.

Contribution

It proposes a novel, disentangled face editing framework using 3D priors and style codes, outperforming prior methods in quality and versatility.

Findings

Achieves state-of-the-art or competitive results across multiple face editing tasks.

Effectively disentangles identity, expression, pose, and illumination.

Demonstrates high-fidelity, controllable face manipulations in real-world scenarios.

Abstract

Face attribute editing aims to generate faces with one or multiple desired face attributes manipulated while other details are preserved. Unlike prior works such as GAN inversion, which has an expensive reverse mapping process, we propose a simple feed-forward network to generate high-fidelity manipulated faces. By simply employing some existing and easy-obtainable prior information, our method can control, transfer, and edit diverse attributes of faces in the wild. The proposed method can consequently be applied to various applications such as face swapping, face relighting, and makeup transfer. In our method, we decouple identity, expression, pose, and illumination using 3D priors; separate texture and colors by using region-wise style codes. All the information is embedded into adversarial learning by our identity-style normalization module. Disentanglement losses are proposed to enhance the generator to extract information independently from each attribute. Comprehensive quantitative and qualitative evaluations have been conducted. In a single framework, our method achieves the best or competitive scores on a variety of face applications.