Face anonymization preserving facial expressions and photometric realism

Abstract

The widespread sharing of face images on social media platforms and in large-scale datasets raises pressing privacy concerns, as biometric identifiers can be exploited without consent. Face anonymization seeks to generate realistic facial images that irreversibly conceal the subject's identity while preserving their usefulness for downstream tasks. However, most existing generative approaches focus on identity removal and image realism, often neglecting facial expressions as well as photometric consistency -- specifically attributes such as illumination and skin tone -- that are critical for applications like relighting, color constancy, and medical or affective analysis. In this work, we propose a feature-preserving anonymization framework that extends DeepPrivacy by incorporating dense facial landmarks to better retain expressions, and by introducing lightweight post-processing modules that ensure consistency in lighting direction and skin color. We further establish evaluation metrics specifically designed to quantify expression fidelity, lighting consistency, and color preservation, complementing standard measures of image realism, pose accuracy, and re-identification resistance. Experiments on the CelebA-HQ dataset demonstrate that our method produces anonymized faces with improved realism and significantly higher fidelity in expression, illumination, and skin tone compared to state-of-the-art baselines. These results underscore the importance of feature-aware anonymization as a step toward more useful, fair, and trustworthy privacy-preserving facial data.

Figures (6)

• Figure 1: Comparison of canonical image processing methods for face anonymization. From left to right are masking, blurring, pixelation, noise, and generation hukkelaas2019deepprivacy.
• Figure 2: Comparison of pose-conditioning landmarks for anonymization. (a) Sparse set of seven landmarks extracted with Mask R-CNN he2017mask, covering only ears, eyes, nose, and shoulders as in the original DeepPrivacy framework. (b) Dense set of 68 facial landmarks detected with the DLib implementation of kazemi2014one, providing finer geometric detail and enabling more faithful expression preservation.
• Figure 3: Relighting pipeline. The original image $I_o$ and anonymized image $I_a$ are decomposed into albedo ($A$) and shading ($S$) components. The shading of $I_o$ is combined with the albedo of $I_a$, and a Laplacian pyramid is used to refine consistency across spatial frequencies, producing the blended output $I_{\text{blend}}$.
• Figure 4: Albedo, normal and shading images estimated for a face image by the SfSNet model sengupta2018sfsnet.
• Figure 5: Aggregated performance across five evaluation dimensions: image quality, privacy, geometry, lighting, and color. Each axis represents the normalized average score for a dimension, where higher values indicate better performance.