TarPro: Targeted Protection against Malicious Image Editing

2025-03-18

AI Summary

To address the security risk of image editing models being misused for generating NSFW content—where existing defenses struggle to simultaneously block malicious edits and preserve fidelity of legitimate edits—this paper proposes a semantic-aware selective protection mechanism. First, a semantic-constrained module is introduced to precisely identify potential adversarial editing intents. Second, a lightweight learnable perturbation generator is designed to apply differentiated interventions specifically on harmful editing pathways. Third, a robustness-optimized training strategy is incorporated to enhance generalization across diverse models and scenarios. Evaluated on multiple state-of-the-art editing models and real-world benchmarks, our method achieves a 96.2% malicious-edit blocking rate while incurring less than 1.3% fidelity degradation on benign edits—outperforming all prior approaches. To our knowledge, this is the first work to enable fine-grained, semantics-driven discrimination and protection of editing behaviors, establishing a new paradigm for secure and controllable image editing.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Security and Privacy: Large-scale security measurements

Abstract

The rapid advancement of image editing techniques has raised concerns about their misuse for generating Not-Safe-for-Work (NSFW) content. This necessitates a targeted protection mechanism that blocks malicious edits while preserving normal editability. However, existing protection methods fail to achieve this balance, as they indiscriminately disrupt all edits while still allowing some harmful content to be generated. To address this, we propose TarPro, a targeted protection framework that prevents malicious edits while maintaining benign modifications. TarPro achieves this through a semantic-aware constraint that only disrupts malicious content and a lightweight perturbation generator that produces a more stable, imperceptible, and robust perturbation for image protection. Extensive experiments demonstrate that TarPro surpasses existing methods, achieving a high protection efficacy while ensuring minimal impact on normal edits. Our results highlight TarPro as a practical solution for secure and controlled image editing.

Problem

• Prevent malicious image edits while preserving normal edits
• Balance between blocking harmful content and maintaining editability
• Develop a robust and imperceptible image protection framework

Innovation

• Semantic-aware constraint blocks malicious edits
• Lightweight perturbation generator ensures robust protection
• Maintains normal editability while preventing harmful content