AI Paper Atlas

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

Towards Robust Model Watermark via Reducing Parametric Vulnerability

Guanhao Gan, Yiming Li, Dongxian Wu, Shu-Tao Xia

ICCV 2023 · Adversarial robustness and security › Attacks on text and graphs · approach: Neural network watermarking · 17 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 adversarial parametric perturbation (APP) alone, without c-BN (ablation)vanilla watermarking · CIFAR-10 Content watermarkaverage WSR drop: 59.15 -> 62.91Hurt
add clean-sample BatchNorm (c-BN) alone, without APP (ablation)vanilla watermarking · CIFAR-10 Content watermarkaverage WSR drop: 59.15 -> 82.24Hurt
combine APP with c-BN (ablation)vanilla watermarking · CIFAR-10 Content watermarkaverage WSR drop: 59.15 -> 10.10Improved
APP + c-BN versus baselines on unrelated-domain watermarkmodel watermark · CIFAR-10 Unrelated watermarkaverage WSR drop: 50.90 -> 6.20Improved
APP + c-BN on ImageNet-subset content watermarkmodel watermark · ImageNet-subset Content watermarkaverage WSR drop: 78.20 -> 39.40Improved
APP + c-BN effect on benign accuracymodel watermark · CIFAR-10benign accuracy: 93.86 -> 93.42No clear effect
APP + c-BN on ImageNet-subset benign accuracymodel watermark · ImageNet-subset Content watermarkbenign accuracy: 74.81 -> 72.29Hurt
apply APP-based method to different target classes (ablation)model watermark · CIFAR-10 Content watermarkaverage WSR under removal attacks: 32.91% -> 94.87% (class 1)Improved
APP versus NNL attack on noise watermarkAPP + c-BN · CIFAR-10 Noise watermarkWSR after NNL: 1.00 -> 3.22No clear effect
vary APP perturbation magnitude within epsilon <= 4.0e-2 (ablation)APP + c-BN · CIFAR-10 Content watermarkrobustness against attacksImproved

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