Attack-Resistant Watermarking for AIGC Image Forensics via Diffusion-based Semantic Deflection
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915720461287424 |
|---|---|
| author | Liu, Qingyu Zhang, Yitao Ba, Zhongjie Shuai, Chao Cheng, Peng Zheng, Tianhang Wang, Zhibo |
| author_facet | Liu, Qingyu Zhang, Yitao Ba, Zhongjie Shuai, Chao Cheng, Peng Zheng, Tianhang Wang, Zhibo |
| contents | Protecting the copyright of user-generated AI images is an emerging challenge as AIGC becomes pervasive in creative workflows. Existing watermarking methods (1) remain vulnerable to real-world adversarial threats, often forced to trade off between defenses against spoofing and removal attacks; and (2) cannot support semantic-level tamper localization. We introduce PAI, a training-free inherent watermarking framework for AIGC copyright protection, plug-and-play with diffusion-based AIGC services. PAI simultaneously provides three key functionalities: robust ownership verification, attack detection, and semantic-level tampering localization. Unlike existing inherent watermark methods that only embed watermarks at noise initialization of diffusion models, we design a novel key-conditioned deflection mechanism that subtly steers the denoising trajectory according to the user key. Such trajectory-level coupling further strengthens the semantic entanglement of identity and content, thereby further enhancing robustness against real-world threats. Moreover, we also provide a theoretical analysis proving that only the valid key can pass verification. Experiments across 12 attack methods show that PAI achieves 98.43\% verification accuracy, improving over SOTA methods by 37.25\% on average, and retains strong tampering localization performance even against advanced AIGC edits. Our code is available at https://github.com/QingyuLiu/PAI. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_06639 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Attack-Resistant Watermarking for AIGC Image Forensics via Diffusion-based Semantic Deflection Liu, Qingyu Zhang, Yitao Ba, Zhongjie Shuai, Chao Cheng, Peng Zheng, Tianhang Wang, Zhibo Cryptography and Security Artificial Intelligence Protecting the copyright of user-generated AI images is an emerging challenge as AIGC becomes pervasive in creative workflows. Existing watermarking methods (1) remain vulnerable to real-world adversarial threats, often forced to trade off between defenses against spoofing and removal attacks; and (2) cannot support semantic-level tamper localization. We introduce PAI, a training-free inherent watermarking framework for AIGC copyright protection, plug-and-play with diffusion-based AIGC services. PAI simultaneously provides three key functionalities: robust ownership verification, attack detection, and semantic-level tampering localization. Unlike existing inherent watermark methods that only embed watermarks at noise initialization of diffusion models, we design a novel key-conditioned deflection mechanism that subtly steers the denoising trajectory according to the user key. Such trajectory-level coupling further strengthens the semantic entanglement of identity and content, thereby further enhancing robustness against real-world threats. Moreover, we also provide a theoretical analysis proving that only the valid key can pass verification. Experiments across 12 attack methods show that PAI achieves 98.43\% verification accuracy, improving over SOTA methods by 37.25\% on average, and retains strong tampering localization performance even against advanced AIGC edits. Our code is available at https://github.com/QingyuLiu/PAI. |
| title | Attack-Resistant Watermarking for AIGC Image Forensics via Diffusion-based Semantic Deflection |
| topic | Cryptography and Security Artificial Intelligence |
| url | https://arxiv.org/abs/2601.06639 |