CoNFies: Controllable Neural Face Avatars

TL;DR

CoNFies introduces a controllable neural face avatar system that uses automated facial action recognition to enable interpretable and high-fidelity expression synthesis from 2D images.

Contribution

It presents a novel framework combining NeRFs with automated facial action recognition for controllable, semantic, and high-quality face avatar synthesis.

Findings

Outperforms existing methods in expression synthesis quality

Provides semantic control over facial expressions using action units

Achieves high anatomical fidelity in synthesized expressions

Abstract

Neural Radiance Fields (NeRF) are compelling techniques for modeling dynamic 3D scenes from 2D image collections. These volumetric representations would be well suited for synthesizing novel facial expressions but for two problems. First, deformable NeRFs are object agnostic and model holistic movement of the scene: they can replay how the motion changes over time, but they cannot alter it in an interpretable way. Second, controllable volumetric representations typically require either time-consuming manual annotations or 3D supervision to provide semantic meaning to the scene. We propose a controllable neural representation for face self-portraits (CoNFies), that solves both of these problems within a common framework, and it can rely on automated processing. We use automated facial action recognition (AFAR) to characterize facial expressions as a combination of action units (AU) and their intensities. AUs provide both the semantic locations and control labels for the system. CoNFies outperformed competing methods for novel view and expression synthesis in terms of visual and anatomic fidelity of expressions.