Fake It Till You Make It: Face analysis in the wild using synthetic data alone

We demonstrate that it is possible to perform face-related computer vision in the wild using synthetic data alone. The community has long enjoyed the benefits of synthesizing training data with graphics, but the domain gap between real and synthetic data has remained a problem, especially for human faces. Researchers have tried to bridge this gap with data mixing, domain adaptation, and domain-adversarial training, but we show that it is possible to synthesize data with minimal domain gap, so that models trained on synthetic data generalize to real in-the-wild datasets. We describe how to combine a procedurally-generated parametric 3D face model with a comprehensive library of hand-crafted assets to render training images with unprecedented realism and diversity. We train machine learning systems for face-related tasks such as landmark localization and face parsing, showing that synthetic data can both match real data in accuracy as well as open up new approaches where manual labelling would be impossible.

PDF Abstract ICCV 2021 PDF ICCV 2021 Abstract

Results from the Paper

Ranked #2 on Face Parsing on Helen (using extra training data)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Face Alignment 300W FakeIt NME_inter-ocular (%, Common) 3.09 # 18
NME_inter-ocular (%, Challenge) 4.86 # 8
Face Parsing Helen UNet (synthetic) Mean F1 92 # 2
Face Parsing Helen UNet (real) Mean F1 91.6 # 3
Face Parsing LaPa UNet (synthetic) Mean F1 90.1 # 7
Face Parsing LaPa UNet (real) Mean F1 90.9 # 6


No methods listed for this paper. Add relevant methods here