Research map / Image generation / Autoregressive and latent generation
MaxMark: High-Capacity Diffusion-Native Watermarking via Robust and Invertible Latent Embedding
Xuanhang Chang, Zhonghao Yang, Cheng Zhuo, Yu Li
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 INN distribution transformation module (ablation) | MaxMark · COCO, 1024 bits | bit accuracy (%): 91.2 -> 99.9 | Improved |
| add INN distribution transformation module (ablation) | MaxMark · COCO, 1024 bits | FID: 388.4 -> 42.0 | Improved |
| add Reed-Solomon ECC with automatic hyperparameter search (ablation) | MaxMark · payload 8192 bits | bit accuracy (%): 94.2 -> 96.2 | Improved |
| use random ECC settings instead of automatic search (ablation) | MaxMark · payload 8192 bits | bit accuracy (%): 96.2 -> 95.1 | Hurt |
| swap Reed-Solomon for BCH (ablation) | MaxMark · payload 4096 bits | bit accuracy (%): 97.7 -> 97.3 | No clear effect |
| embed in sign bit rather than low-order bits (ablation) | MaxMark · fp32 latent bit positions | bit accuracy | Improved |
| increase payload from 4096 to 16384 bits in baselines | PRC Watermark · SD v1.5 | bit accuracy (%): 50.1 -> 43.6 (4096 to 16384) | Hurt |
| apply to video latents with distribution transformation | MaxMark · ModelScope 32k bits | bit accuracy (%): 54.6 -> 98.2 vs PRC | Improved |
| apply to audio latents with distribution transformation | MaxMark · AudioLDM2 32k bits | bit accuracy (%): 48.8 -> 96.7 vs PRC | Improved |
Closest papers by meaning
- T2SMark: Balancing Robustness and Diversity in Noise-as-Watermark for Diffusion Models (NeurIPS 2025 · similarity 0.90)
- SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models (CVPR 2025 · similarity 0.90)
- Shallow Diffuse: Robust and Invisible Watermarking through Low-Dim Subspaces in Diffusion Models (NeurIPS 2025 · similarity 0.90)
- WMAdapter: Adding WaterMark Control to Latent Diffusion Models (ICML 2025 · similarity 0.90)
- StableGuard: Towards Unified Copyright Protection and Tamper Localization in Latent Diffusion Models (NeurIPS 2025 · similarity 0.89)
- ROBIN: Robust and Invisible Watermarks for Diffusion Models with Adversarial Optimization (NeurIPS 2024 · similarity 0.89)
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.
