Research map / Adversarial robustness and security / Attacks on text and graphs
Adversarial Camouflage: Hiding Physical-World Attacks With Natural Styles
Check 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 |
|---|---|---|---|
| Use unrestricted style-based camouflage instead of PGD-16's bounded perturbation | AdvCam · Human perception study | natural and realistic selections: 77.3 ± 1.53% -> 80.7 ± 1.53% | No clear effect |
| Use style-based camouflage instead of an adversarial patch | AdvCam · Human perception study | natural and realistic selections: 19.0 ± 1.68% -> 80.7 ± 1.53% | Helped |
| Increase adversarial strength parameter λ (ablation) | AdvCam · ImageNet, randomly selected 2000 test images | attack success rate: by up to 20% | Helped |
| Increase attack region size from 40*40 to 120*120 (ablation) | AdvCam · ImageNet, randomly selected 2000 test images | attack success rate: by up to 40% | Helped |
| Add content preservation loss (ablation) | AdvCam · ImageNet camouflage examples | preservation of original content | Helped |
| Add smoothness enhancement loss (ablation) | AdvCam · ImageNet camouflage examples | smoothness of object surface | Helped |
| Add content preservation loss (ablation) | AdvCam · ImageNet table lamp example | visual appearance | No clear effect |
| Add smoothness enhancement loss (ablation) | AdvCam · ImageNet camouflage examples | visual appearance | No clear effect |
Closest papers by meaning
- Dual Attention Suppression Attack: Generate Adversarial Camouflage in Physical World (CVPR 2021 · similarity 0.91)
- Unnoticed Yet Effective: A Hybrid Physical Camouflage Framework Against DNNs and Human Perception (AAAI 2026 · similarity 0.91)
- PhyCamo: A Robust Physical Camouflage via Contrastive Learning for Multi-View Physical Adversarial Attack (AAAI 2025 · similarity 0.89)
- CNCA: Toward Customizable and Natural Generation of Adversarial Camouflage for Vehicle Detectors (NeurIPS 2024 · similarity 0.87)
- Breaking Barriers in Physical-World Adversarial Examples: Improving Robustness and Transferability via Robust Feature (AAAI 2025 · similarity 0.87)
- RAUCA: A Novel Physical Adversarial Attack on Vehicle Detectors via Robust and Accurate Camouflage Generation (ICML 2024 · similarity 0.86)
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.