JPEG Compressed Images Can Bypass Protections Against AI Editing

TL;DR

This paper reveals that JPEG compression can bypass existing imperceptible perturbation protections against AI image editing, highlighting the need for more robust defense methods.

Contribution

It demonstrates the vulnerability of current perturbation-based protections to JPEG compression and emphasizes the importance of developing more robust image protection techniques.

Findings

JPEG compression can nullify existing protections

Imperceptible perturbations are not robust to common image formats

Calls for alternative, more resilient protection methods

Abstract

Recently developed text-to-image diffusion models make it easy to edit or create high-quality images. Their ease of use has raised concerns about the potential for malicious editing or deepfake creation. Imperceptible perturbations have been proposed as a means of protecting images from malicious editing by preventing diffusion models from generating realistic images. However, we find that the aforementioned perturbations are not robust to JPEG compression, which poses a major weakness because of the common usage and availability of JPEG. We discuss the importance of robustness for additive imperceptible perturbations and encourage alternative approaches to protect images against editing.