Research map / 3D vision / NeRF and radiance fields
CopyRNeRF: Protecting the CopyRight of Neural Radiance Fields
Ziyuan Luo, Qing Guo, Ka Chun Cheung, Simon See, Renjie Wan
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 |
|---|---|---|---|
| embed message in color representation rather than 2D watermarking before NeRF training | NeRF · Blender+LLFF, 16 bits | bit accuracy: 50.19% -> 91.16% (vs HiDDeN+NeRF) | Improved |
| embed message in color representation instead of geometry (ablation) | CopyRNeRF · Blender+LLFF, 16 bits | bit accuracy: 60.16% -> 91.16% | Improved |
| add distortion-resistant rendering (DRR) (ablation) | CopyRNeRF · Rotation, 16 bits | bit accuracy: 75.81% -> 88.13% | Improved |
| add distortion-resistant rendering (DRR) (ablation) | CopyRNeRF · no distortion, 16 bits | bit accuracy: 91.25% -> 91.16% | No clear effect |
| add message feature field (ablation) | CopyRNeRF · feature field ablation | PSNR / bit accuracy: PSNR 20.46 -> 32.68; bit acc 82.69% -> 100% | Improved |
| add color feature field (ablation) | CopyRNeRF · feature field ablation | PSNR / bit accuracy: PSNR 21.06 -> 32.68; bit acc 80.69% -> 100% | Improved |
| concatenate message to NeRF input instead of watermarked color field | NeRF · Blender+LLFF, 16 bits | bit accuracy: 91.16% -> 52.22% | Hurt |
| increase message length | CopyRNeRF · Blender+LLFF | bit accuracy | Hurt |
| use different rendering sampling scheme at test time than training (ablation) | CopyRNeRF · Blender+LLFF | bit accuracy / PSNR: PSNR 32.68 -> 32.17; bit acc 100% -> 100% | No clear effect |
Later papers that compared with this method
4 later accepted papers report CopyRNeRF as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- The NeRF Signature: Codebook-Aided Watermarking for Neural Radiance Fields (TPAMI 2025 · better on 5 of 5 shared results)
- WateRF: Robust Watermarks in Radiance Fields for Protection of Copyrights (CVPR 2024 · better on 6 of 6 shared results)
- Protecting NeRFs' Copyright via Plug-And-Play Watermarking Base Model (ECCV 2024 · better on 4 of 5 shared results)
- GeometrySticker: Enabling Ownership Claim of Recolorized Neural Radiance Fields (ECCV 2024 · 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
- Protecting NeRFs' Copyright via Plug-And-Play Watermarking Base Model (ECCV 2024 · similarity 0.95)
- WateRF: Robust Watermarks in Radiance Fields for Protection of Copyrights (CVPR 2024 · similarity 0.92)
- GeometrySticker: Enabling Ownership Claim of Recolorized Neural Radiance Fields (ECCV 2024 · similarity 0.91)
- The NeRF Signature: Codebook-Aided Watermarking for Neural Radiance Fields (TPAMI 2025 · similarity 0.91)
- DreaMark: Rooting Watermark in Score Distillation Sampling Generated Neural Radiance Fields (AAAI 2025 · similarity 0.88)
- Aleth-NeRF: Illumination Adaptive NeRF with Concealing Field Assumption (AAAI 2024 · similarity 0.86)
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.
