On Learning 3D Face Morphable Model from In-the-wild Images

TL;DR

This paper introduces a nonlinear 3D face morphable model learned from in-the-wild images, overcoming limitations of traditional models that require 3D scans, and demonstrates improved face analysis capabilities.

Contribution

It proposes a novel end-to-end trainable framework to learn a nonlinear 3DMM directly from in-the-wild images without 3D scans, enhancing representation power.

Findings

Outperforms linear 3DMM in face alignment and 3D reconstruction

Enables face editing with improved accuracy

Demonstrates effective learning from large-scale in-the-wild data

Abstract

As a classic statistical model of 3D facial shape and albedo, 3D Morphable Model (3DMM) is widely used in facial analysis, e.g., model fitting, image synthesis. Conventional 3DMM is learned from a set of 3D face scans with associated well-controlled 2D face images, and represented by two sets of PCA basis functions. Due to the type and amount of training data, as well as, the linear bases, the representation power of 3DMM can be limited. To address these problems, this paper proposes an innovative framework to learn a nonlinear 3DMM model from a large set of in-the-wild face images, without collecting 3D face scans. Specifically, given a face image as input, a network encoder estimates the projection, lighting, shape and albedo parameters. Two decoders serve as the nonlinear 3DMM to map from the shape and albedo parameters to the 3D shape and albedo, respectively. With the projection parameter, lighting, 3D shape, and albedo, a novel analytically-differentiable rendering layer is designed to reconstruct the original input face. The entire network is end-to-end trainable with only weak supervision. We demonstrate the superior representation power of our nonlinear 3DMM over its linear counterpart, and its contribution to face alignment, 3D reconstruction, and face editing.