Machine Pareidolia: Protecting Facial Image with Emotional Editing

TL;DR

This paper introduces MAP, a novel facial privacy protection method that uses emotional editing to disguise identities, outperforming previous techniques in preserving privacy and image quality across diverse scenarios.

Contribution

The paper presents a new emotion-based identity disguise technique that improves transferability and applicability over existing makeup and noise-based methods.

Findings

Outperforms previous methods in qualitative and quantitative evaluations.

Effective against online facial recognition APIs.

Maintains high perceptual quality in protected images.

Abstract

The proliferation of facial recognition (FR) systems has raised privacy concerns in the digital realm, as malicious uses of FR models pose a significant threat. Traditional countermeasures, such as makeup style transfer, have suffered from low transferability in black-box settings and limited applicability across various demographic groups, including males and individuals with darker skin tones. To address these challenges, we introduce a novel facial privacy protection method, dubbed \textbf{MAP}, a pioneering approach that employs human emotion modifications to disguise original identities as target identities in facial images. Our method uniquely fine-tunes a score network to learn dual objectives, target identity and human expression, which are jointly optimized through gradient projection to ensure convergence at a shared local optimum. Additionally, we enhance the perceptual quality of protected images by applying local smoothness regularization and optimizing the score matching loss within our network. Empirical experiments demonstrate that our innovative approach surpasses previous baselines, including noise-based, makeup-based, and freeform attribute methods, in both qualitative fidelity and quantitative metrics. Furthermore, MAP proves its effectiveness against an online FR API and shows advanced adaptability in uncommon photographic scenarios.