Learning Complete 3D Morphable Face Models from Images and Videos

TL;DR

This paper introduces a novel self-supervised method to learn complete 3D face models, including identity, expression, and albedo, from images and videos, improving generalization and reconstruction quality.

Contribution

It is the first to learn fully disentangled 3D face models from in-the-wild data, including an expression basis, using a self-supervised approach.

Findings

Models generalize better across diverse identities and expressions.

Achieves higher quality monocular face reconstructions.

Enables in-the-wild 3D face modeling without 3D scans.

Abstract

Most 3D face reconstruction methods rely on 3D morphable models, which disentangle the space of facial deformations into identity geometry, expressions and skin reflectance. These models are typically learned from a limited number of 3D scans and thus do not generalize well across different identities and expressions. We present the first approach to learn complete 3D models of face identity geometry, albedo and expression just from images and videos. The virtually endless collection of such data, in combination with our self-supervised learning-based approach allows for learning face models that generalize beyond the span of existing approaches. Our network design and loss functions ensure a disentangled parameterization of not only identity and albedo, but also, for the first time, an expression basis. Our method also allows for in-the-wild monocular reconstruction at test time. We show that our learned models better generalize and lead to higher quality image-based reconstructions than existing approaches.