Research map / 3D vision / Gaussian splatting
Mark4D: Temporally-Consistent Watermarking for 4D Gaussian Splatting
Jaejin Lee, Minjae Jeong, Joonhyuk Park, Yechan Hwang, Seunghun Baek, Won Hwa Kim
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add delta_i(t) SH offsets to position-only offsets (ablation) | Mark4D · D-NeRF+DyNeRF avg, 48 bits | Bit Acc (%): 79.55 -> 91.27 bit acc; PSNR 30.96 -> 33.94 | Mixed |
| add position offsets to SH-only offsets (unconstrained) (ablation) | Mark4D · D-NeRF+DyNeRF avg, 48 bits | PSNR: 37.82 -> 33.94 | Hurt |
| add trajectory alignment loss L_align (ablation) | Mark4D · D-NeRF+DyNeRF avg, 48 bits | Bit Acc (%): 91.27 -> 92.94; PSNR 33.94 -> 39.88 | Improved |
| add motion-adaptive loss weighting (ablation) | Mark4D · D-NeRF+DyNeRF avg, 48 bits | Bit Acc (%): 92.94 -> 95.02; PSNR 39.88 -> 41.45 | Improved |
| add motion-adaptive weighting (ablation) | Mark4D · DyNeRF | Bit Acc (%): 93.41 -> 96.55 (+3.14) | Improved |
| add motion-adaptive weighting (ablation) | Mark4D · D-NeRF | Bit Acc (%): 92.59 -> 93.88 (+1.29) | Improved |
| add motion-adaptive weighting (ablation) | Mark4D · DyNeRF (flame salmon) | temporal stability of bit accuracy and PSNR | Improved |
| use X-CLIP latent-space decoder with trajectory offsets instead of 3D-GSW at 64 bits | 4DGS watermarking · D-NeRF+DyNeRF avg, 64 bits | Bit Acc (%): 83.36 -> 92.71 | Improved |
| increase message length from 32 to 64 bits | Mark4D · D-NeRF+DyNeRF avg | Bit Acc (%): 96.34 -> 92.71 | Hurt |
| use video-level Mark4D vs VideoSeal under JPEG compression | 4DGS watermarking · D-NeRF+DyNeRF avg, 32 bits | Bit Acc (%): 74.28 -> 91.93 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- 3D-GSW — 3D-GSW: 3D Gaussian Splatting for Robust Watermarking (CVPR 2025) · better on 5 of 5 shared results
- GuardSplat — GuardSplat: Efficient and Robust Watermarking for 3D Gaussian Splatting (CVPR 2025) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Where, What, Why: Toward Explainable 3D-GS Watermarking (CVPR 2026 · similarity 0.90)
- Write Where It Matters: Policy-Guided Watermarks for 3D Gaussian Splatting (CVPR 2026 · similarity 0.90)
- GaussianMarker: Uncertainty-Aware Copyright Protection of 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.89)
- GuardSplat: Efficient and Robust Watermarking for 3D Gaussian Splatting (CVPR 2025 · similarity 0.88)
- 3D-GSW: 3D Gaussian Splatting for Robust Watermarking (CVPR 2025 · similarity 0.87)
- Gaussian Shading: Provable Performance-Lossless Image Watermarking for Diffusion Models (CVPR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
