arXiv cs.AI· Sifan Zhou, Hang Zhang, Yuhang Wang, Ming Li·· 9 天前AI 评分22
4D-GSW:面向 4D Gaussian Splatting 的运动学感知时空一致水印框架
4D-GSW: Kinematic-Aware Spatio-Temporal Consistent Watermarking for 4D Gaussian Splatting
AI 导读
针对 4D Gaussian Splatting(4DGS)资产版权保护问题,研究者提出运动学感知水印框架 4D-GSW,可在嵌入版权信息的同时保持高时空一致性。
正文 · 原文
Abstract:While 4D Gaussian Splatting (4DGS) has revolutionized high-fidelity dynamic reconstruction, safeguarding the intellectual property of these assets remains an open challenge. Conventional steganographic techniques often neglect the underlying kinematic manifolds, triggering non-physical artifacts such as severe temporal flickering and "FVD collapse". To address this, we propose \textbf{4D-GSW}, a kinematic-aware watermarking framework designed to embed robust copyright information while preserving high spatio-temporal consistency. Unlike prior 4D steganography that primarily focuses on opacity-guided invisibility, our approach explicitly addresses the physical coherence of motion trajectories. We introduce a \textbf{Spatio-Temporal Curvature (STC)} metric to identify "Dynamic Instants," adaptively gating watermark gradient injection to shield critical motion manifolds from non-physical perturbations. To ensure global coherence across complex deformations, we formulate a joint \textbf{HMM-MRF energy minimization} model that synchronizes watermark phases within both temporal trajectories and spatial neighborhoods. Furthermore, an \textbf{anisotropic gradient routing} mechanism ensures that watermark embedding remains strictly decoupled from photometric reconstruction fidelity. Extensive experiments have demonstrated the superior performance of our method in robustly hiding watermarks while resisting various attacks and maintaining high rendering quality and spatiotemporal consistency.
| Comments: | 9 pages main paper, 7 figures, 18 pages in total |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.22342 [cs.CV] |
| (or arXiv:2605.22342v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2605.22342 arXiv-issued DOI via DataCite |
Submission history
From: Sifan Zhou [view email]
[v1]
Thu, 21 May 2026 11:27:11 UTC (2,726 KB)
[v2]
Wed, 7 Oct 2026 17:20:31 UTC (2,758 KB)
来源:arXiv cs.AI · arxiv.org