AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

A Secure Image Watermarking Framework with Statistical Guarantees via Adversarial Attacks on Secret Key Networks

Feiyu CHEN, Wei Lin, Ziquan Liu, Antoni Chan

ECCV 2024 · Adversarial robustness and security › Attacks on text and graphs · approach: Neural network watermarking · 2 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
use a secret, uniquely trained key network instead of a publicly available pre-trained CNNzero-bit watermarking · Model-fake attackfalse rate (%): 100.00% -> 4.00%Improved
use the secret key network against recursive overwriting with 5 watermarkszero-bit watermarking · Overlapping watermarksdetection rate of original watermark: 12.00% -> 99.60% vs SSLWM (Ours(S))Improved
combine length and angle hypothesis tests (ablation)watermark detector · watermarked images, alpha 0.05percent of p-values below 0.05: 95.74% -> 99.48% (length only vs combined)Improved
use L2-norm PGD with gradient value instead of sign and L-infinity (ablation)watermark embedding attack · robustness to distortionssuccess detection rateImproved
use single-step FGSM-G instead of iterative PGD (ablation)watermark embedding attack · robustness to distortionssuccess detection rateHurt
add a target length to the adversarial loss (ablation)watermark embedding · robustness to distortionssuccess detection rateImproved
use angle-only adversarial loss instead of length plus angle (ablation)watermark embedding · robustness to distortionssuccess detection rateHurt
add data augmentation (rotation and cropping) during watermark embedding (ablation)watermark embedding · robustness to distortionssuccess detection rateImproved
use the SKN approach versus baselines under rotation and croppingzero-bit watermarking · rotation and croppingAvg DR: cropping 97.90% (DNN0B) vs 90.58%; rotation 86.43% (SSLWM) vs 85.89%Hurt
use the SKN approach instead of SSLWM at equal PSNR 32watermarking imperceptibility · MSCOCO testSSIM: 0.9026 -> 0.9768Improved

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