Label-guided Facial Retouching Reversion

TL;DR

This paper introduces Re-Face, a framework for reversing facial retouching in images by detecting retouching, reverting it with a guided model, and correcting color shifts, improving photo authenticity.

Contribution

The paper presents a novel three-component framework, including a retouching detector, a guided reversion model, and a color correction module, specifically addressing geometric deformations caused by retouching.

Findings

Re-Face effectively reverts retouched facial images.

The framework accurately detects retouching and corrects color shifts.

Experiments show significant improvements over existing methods.

Abstract

With the popularity of social media platforms and retouching tools, more people are beautifying their facial photos, posing challenges for fields requiring photo authenticity. To address this issue, some work has proposed makeup removal methods, but they cannot revert images involving geometric deformations caused by retouching. To tackle the problem of facial retouching reversion, we propose a framework, dubbed Re-Face, which consists of three components: a facial retouching detector, an image reversion model named FaceR, and a color correction module called Hierarchical Adaptive Instance Normalization (H-AdaIN). FaceR can utilize labels generated by the facial retouching detector as guidance to revert the retouched facial images. Then, color correction is performed using H-AdaIN to address the issue of color shift. Extensive experiments demonstrate the effectiveness of our framework and each module.