Universal Anti-forensics Attack against Image Forgery Detection via Multi-modal Guidance

TL;DR

This paper introduces ForgeryEraser, a universal anti-forensics attack method that exploits vulnerabilities in vision-language models to deceive and degrade the performance of state-of-the-art image forgery detectors.

Contribution

The paper presents a novel attack framework that does not require access to target detectors, leveraging multi-modal guidance to erase forgery traces in images.

Findings

ForgeryEraser significantly reduces detection accuracy of advanced AIGC detectors.

The attack causes forged images to produce explanations similar to authentic images.

The method exploits systemic reliance on shared vision-language model backbones.

Abstract

The rapid advancement of AI-Generated Content (AIGC) technologies poses significant challenges for authenticity assessment. However, existing evaluation protocols largely overlook anti-forensics attack, failing to ensure the comprehensive robustness of state-of-the-art AIGC detectors in real-world applications. To bridge this gap, we propose ForgeryEraser, a framework designed to execute universal anti-forensics attack without access to the target AIGC detectors. We reveal an adversarial vulnerability stemming from the systemic reliance on Vision-Language Models (VLMs) as shared backbones (e.g., CLIP), where downstream AIGC detectors inherit the feature space of these publicly accessible models. Instead of traditional logit-based optimization, we design a multi-modal guidance loss to drive forged image embeddings within the VLM feature space toward text-derived authentic anchors to erase forgery traces, while repelling them from forgery anchors. Extensive experiments demonstrate that ForgeryEraser causes substantial performance degradation to advanced AIGC detectors on both global synthesis and local editing benchmarks. Moreover, ForgeryEraser induces explainable forensic models to generate explanations consistent with authentic images for forged images. Our code will be made publicly available.