AuraMask: An Extensible Pipeline for Developing Aesthetic Anti-Facial Recognition Image Filters

2026-05-12

AI Summary

Existing adversarial face recognition filters suffer from severe visual distortion that undermines users’ self-expression, leading to low adoption rates. This work proposes AuraMask, the first approach to systematically integrate aesthetic design into adversarial filter generation by emulating popular Instagram styles, thereby effectively evading facial recognition systems while preserving high visual appeal. Combining adversarial example generation, style transfer, and human-centered evaluation, AuraMask establishes an end-to-end, scalable pipeline for filter creation and validation. Experimental results demonstrate that the 40 filters generated by AuraMask achieve or surpass the adversarial performance of state-of-the-art methods on open-source face recognition models. Furthermore, a user study involving 630 participants confirms that AuraMask filters are significantly more acceptable to users than existing alternatives.

Abstract

Anti-facial recognition (AFR) image filters alter images in ways that are subtle to people but blinding to computer vision. Yet, despite widespread interest in these technologies to subvert surveillance, users rarely use them in practice -- because the ``subtle'' alterations are visible enough to conflict with users' self-presentation goals. To address this challenge, we propose AuraMask: a novel approach to creating AFR filters that are both adversarially effective and aesthetically acceptable. Using AuraMask, we produce 40 ``aesthetic'' filters that emulate popular ``one-click'' Instagram image filters. We show that AuraMask filters meet or exceed the adversarial effectiveness of prior methods against open-source facial recognition models. Moreover, in a controlled online user study ($N=630$) we confirm these filters achieve significantly higher user acceptance than prior methods. Lastly, we provide our AFR pipeline to the community for accelerated research in adversarially effective and aesthetically acceptable protections.

Problem

• anti-facial recognition
• image filters
• user acceptance
• aesthetic
• self-presentation

Innovation

• adversarial filters
• aesthetic acceptability
• anti-facial recognition
• image perturbation
• user-centered design