Research map / Image generation / Autoregressive and latent generation
NoisePrints: Distortion-Free Watermarks for Authorship in Private Diffusion Models
Nir Goren, Oren Katzir, Abhinav Nakarmi, Eyal Ronen, Mahmood Sharif, Or Patashnik
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 |
|---|---|---|---|
| replace inversion-based verification with cosine similarity on seed noise | NoisePrints vs WIND/PRC/GS · End-to-end verification | speedup: x14-x213 | Improved |
| replace inversion with cosine similarity | NoisePrints · Wan2.1-1.3B verification time | time: 91.473 s -> 0.097 ms | Improved |
| use seed-noise correlation instead of WIND inversion under DDIM inversion attack | NoisePrints · Traceability SD2.0 | detection rate: 0.560 -> 0.980 | Improved |
| use seed-noise correlation instead of WIND under SDEdit | NoisePrints · Traceability SD2.0 | detection rate: 0.980 -> 1.000 | Improved |
| use seed-noise correlation instead of WIND under Gaussian blur r=4 | NoisePrints · Traceability SD2.0 | detection rate: 1.000 -> 0.990 | No clear effect |
| decode and re-encode latents through the VAE (ablation) | NoisePrints · SD2.0 cosine similarity | mean cosine similarity: 0.4922 -> 0.4818 | No clear effect |
| decode and re-encode latents through the VAE (ablation) | NoisePrints · SDXL cosine similarity | mean cosine similarity: 0.4545 -> 0.4283 | No clear effect |
| re-align geometrically transformed images with estimated inverse transform | NoisePrints · SD2.0 rotation and crop | pass rate: 100% | Improved |
| use low-entropy prompt with large uniform regions | NoisePrints · SDXL and Flux.1-schnell | correlation vs threshold | Hurt |
Closest papers by meaning
- Secure Seed-Based Multi-bit Watermarking for Diffusion Models from First Principles (NeurIPS 2026 · similarity 0.89)
- SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models (CVPR 2025 · similarity 0.88)
- T2SMark: Balancing Robustness and Diversity in Noise-as-Watermark for Diffusion Models (NeurIPS 2025 · similarity 0.88)
- Harnessing Frequency Spectrum Insights for Image Copyright Protection Against Diffusion Models (CVPR 2025 · similarity 0.88)
- Hidden in the Noise: Two-Stage Robust Watermarking for Images (ICLR 2025 · similarity 0.88)
- Break the Trade-off Between Watermark Strength and Speculative Sampling Efficiency for Language Models (ICLR 2026 · similarity 0.88)
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.
