AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Semantic Watermarking Reinvented: Enhancing Robustness and Generation Quality with Fourier Integrity

Sung Ju Lee, Nam Ik Cho

ICCV 2025 · Image generation › Autoregressive and latent generation · approach: Robust diffusion watermarking · 8 citations (OpenAlex)

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 changedWhereMeasuredEffect
add Hermitian symmetry (SFW) with real-only detection (ablation)Tree-Ring · MS-COCO identificationaccuracy: 0.416 -> 0.775Improved
add Hermitian symmetry (SFW) with real-only detection (ablation)Tree-Ring · MS-COCO verificationTPR@1%FPR: 0.805 -> 0.936Improved
add imaginary component to detection once SFW preserves frequency integrity (ablation)HSTR · MS-COCO identificationaccuracy: 0.775 -> 0.889Improved
use both real and imaginary detection without SFW instead of real only (ablation)Tree-Ring · MS-COCO identificationaccuracy: 0.416 -> 0.114Hurt
add center-aware embedding on top of SFW (ablation)HSTR · MS-COCO cropping attacksidentification accuracy: 0.374 -> 1.000Improved
add center-aware embedding on top of SFW (ablation)HSTR · MS-COCO signal processing attacksidentification accuracy: 0.856 -> 0.838Hurt
add center-aware embedding on top of SFW (ablation)HSTR · MS-COCO regeneration attacksidentification accuracy: 0.812 -> 0.878Improved
add SFW to Tree-Ring (no center-aware) (ablation)Tree-Ring · MS-COCOFID: 26.418 -> 25.071Improved
add SFW and center-aware embedding to Tree-Ring (average) (ablation)Tree-Ring · MS-COCO identificationaverage accuracy: 0.114 -> 0.889Improved
add SFW to Tree-Ring latent noiseTree-Ring · latent normality, 1,000 samplesKS failure rate: 0.234 -> 0.071Improved
use HSQR instead of RingID high-energy ring patternHSQR · MS-COCO generation qualityFID: 27.052 -> 24.895Improved
increase diffusion regeneration noise step from 20 to 180 (ablation)HSTR · MS-COCO identificationaccuracy: 0.999 -> 0.781Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy