AI Paper Atlas

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

Exploring Adversarial Fake Images on Face Manifold

Dongze Li, Wei Wang, Hongxing Fan, Jing Dong

CVPR 2021 · Adversarial robustness and security › Attacks on text and graphs · approach: Diffusion-based adversarial purification · 40 citations (OpenAlex)

Check 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
Search adversarial points by updating Style-GAN latent or noise inputs instead of adding pixel-space perturbationsStyle-GAN fake faces against EfficientNet and Xception · White-box attackaccuracy: 97% -> 0%; 93% -> 0%Improved
Use PGD L2 rather than manifold searchEfficientNet and Xception · White-box attackaccuracy: 60%; 63%Hurt
Generate against EfficientNet and test transfer to XceptionStyle-GAN fake faces · Black-box attackaccuracy: 93% -> 50%Improved
Generate against Xception and test transfer to EfficientNetStyle-GAN fake faces · Black-box attackaccuracy: 97% -> 90%No clear effect
Combine model losses, fuse model logits, or alternate updates across modelsEfficientNet and Xception · Ensemble attackaccuracy: 0.5%; 0.1%; 0% on EfficientNet and 0% on XceptionImproved
Add updates to higher-level Style-GAN noise inputs rather than lower-level inputs (ablation)Single-level noise attack · EfficientNet and Xceptionaccuracy: 97% -> 1%; 93% -> 0% (level 0 to levels 6-8)Improved
Update only noise levels 1-3 rather than leaving the image unperturbed (ablation)Single-level noise attack · EfficientNet and Xceptionaccuracy: 97% -> 99%, 98%, 98%; 93% -> 87%, 88%, 85%No clear effect
Update only noise level 4 or 5 (ablation)Single-level noise attack · EfficientNet and Xceptionaccuracy: 75%; 57%; 35%; 23%Improved
Update only noise level 6, 7, or 8 (ablation)Single-level noise attack · EfficientNet and Xceptionaccuracy: 1%; 0%Improved
Use Style-GAN manifold search instead of PGD L_infGenerated adversarial faces · Image quality assessmentLPIPS: 1.084 -> 0.123Improved

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