Research map / Adversarial robustness and security / Attacks on text and graphs
Adv-Attribute: Inconspicuous and Transferable Adversarial Attack on Face Recognition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| encode adversarial noise in facial attribute latents instead of pixel-level perturbations | Adv-Attribute · FFHQ and CelebA-HQ, black-box basic face recognition models | ASR | Helped |
| encode adversarial noise in facial attribute latents instead of pixel-level perturbations | Adv-Attribute · FFHQ and CelebA-HQ, robust face recognition models | ASR | Helped |
| add importance-aware attribute selection (ablation) | Adv-Attribute · FFHQ, IR152 | ASR: 44.10 -> 46.30 | Helped |
| add importance-aware attribute selection (ablation) | Adv-Attribute · FFHQ, MobileFace | ASR: 49.10 -> 49.90 | Helped |
| add importance-aware attribute selection (ablation) | Adv-Attribute · FFHQ, FaceNet | ASR: 30.70 -> 31.90 | Helped |
| add importance-aware attribute selection (ablation) | Adv-Attribute · CelebA-HQ, IR152 | ASR: 43.70 -> 44.30 | Helped |
| add importance-aware attribute selection (ablation) | Adv-Attribute · CelebA-HQ, MobileFace | ASR: 46.40 -> 50.20 | Helped |
| add importance-aware attribute selection (ablation) | Adv-Attribute · CelebA-HQ, FaceNet | ASR: 28.60 -> 31.80 | Helped |
| add multi-objective optimization (ablation) | Adv-Attribute · FFHQ, IR152 | ASR: 45.50 -> 46.30 | Helped |
| add multi-objective optimization (ablation) | Adv-Attribute · FFHQ, MobileFace | ASR: 49.10 -> 49.90 | Helped |
| add multi-objective optimization (ablation) | Adv-Attribute · FFHQ, FaceNet | ASR: 31.30 -> 31.90 | Helped |
| add multi-objective optimization (ablation) | Adv-Attribute · CelebA-HQ, IR152 | ASR: 44.00 -> 44.30 | Helped |
Closest papers by meaning
- Attacks Meet Interpretability: Attribute-steered Detection of Adversarial Samples (NeurIPS 2018 · similarity 0.88)
- Improving the Transferability of Adversarial Attacks on Face Recognition with Diverse Parameters Augmentation (CVPR 2025 · similarity 0.86)
- Transferable Adversarial Face Attack with Text Controlled Attribute (AAAI 2025 · similarity 0.86)
- Pixel2Feature Attack (P2FA): Rethinking the Perturbed Space to Enhance Adversarial Transferability (ICML 2025 · similarity 0.86)
- Non-Adaptive Adversarial Face Generation (NeurIPS 2025 · similarity 0.85)
- Adv-Diffusion: Imperceptible Adversarial Face Identity Attack via Latent Diffusion Model (AAAI 2024 · similarity 0.85)
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.