SHIFT: Stochastic Hidden-Trajectory Deflection for Removing Diffusion-based Watermark
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914436614193152 |
|---|---|
| author | Bao, Rui Gao, Zheng Li, Xiaoyu Feng, Xiaoyan Song, Yang Jiang, Jiaojiao |
| author_facet | Bao, Rui Gao, Zheng Li, Xiaoyu Feng, Xiaoyan Song, Yang Jiang, Jiaojiao |
| contents | Diffusion-based watermarking methods embed verifiable marks by manipulating the initial noise or the reverse diffusion trajectory. However, these methods share a critical assumption: verification can succeed only if the diffusion trajectory can be faithfully reconstructed. This reliance on trajectory recovery constitutes a fundamental and exploitable vulnerability. We propose $\underline{\mathbf{S}}$tochastic $\underline{\mathbf{Hi}}$dden-Trajectory De$\underline{\mathbf{f}}$lec$\underline{\mathbf{t}}$ion ($\mathbf{SHIFT}$), a training-free attack that exploits this common weakness across diverse watermarking paradigms. SHIFT leverages stochastic diffusion resampling to deflect the generative trajectory in latent space, making the reconstructed image statistically decoupled from the original watermark-embedded trajectory while preserving strong visual quality and semantic consistency. Extensive experiments on nine representative watermarking methods spanning noise-space, frequency-domain, and optimization-based paradigms show that SHIFT achieves 95%--100% attack success rates with nearly no loss in semantic quality, without requiring any watermark-specific knowledge or model retraining. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_29742 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SHIFT: Stochastic Hidden-Trajectory Deflection for Removing Diffusion-based Watermark Bao, Rui Gao, Zheng Li, Xiaoyu Feng, Xiaoyan Song, Yang Jiang, Jiaojiao Computer Vision and Pattern Recognition Cryptography and Security Diffusion-based watermarking methods embed verifiable marks by manipulating the initial noise or the reverse diffusion trajectory. However, these methods share a critical assumption: verification can succeed only if the diffusion trajectory can be faithfully reconstructed. This reliance on trajectory recovery constitutes a fundamental and exploitable vulnerability. We propose $\underline{\mathbf{S}}$tochastic $\underline{\mathbf{Hi}}$dden-Trajectory De$\underline{\mathbf{f}}$lec$\underline{\mathbf{t}}$ion ($\mathbf{SHIFT}$), a training-free attack that exploits this common weakness across diverse watermarking paradigms. SHIFT leverages stochastic diffusion resampling to deflect the generative trajectory in latent space, making the reconstructed image statistically decoupled from the original watermark-embedded trajectory while preserving strong visual quality and semantic consistency. Extensive experiments on nine representative watermarking methods spanning noise-space, frequency-domain, and optimization-based paradigms show that SHIFT achieves 95%--100% attack success rates with nearly no loss in semantic quality, without requiring any watermark-specific knowledge or model retraining. |
| title | SHIFT: Stochastic Hidden-Trajectory Deflection for Removing Diffusion-based Watermark |
| topic | Computer Vision and Pattern Recognition Cryptography and Security |
| url | https://arxiv.org/abs/2603.29742 |