Research map / Adversarial robustness and security / Attacks on text and graphs
Physical Adversarial Clothing Evades Visible-Thermal Detectors via Non-Overlapping RGB-T Pattern
Xiaopei Zhu, Guanning Zeng, Zhanhao Hu, Jun Zhu, Xiaolin Hu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add Spatially-Random Discretization to pattern optimization (ablation) | SDCO · Digital world, Prob-L | ASR (%): 67.2 -> 99.8 | Improved |
| add Spatially-Random Discretization to pattern optimization (ablation) | SDCO · Digital world, YOLOv11 T | ASR (%): 46.4 -> 99.4 | Improved |
| add Spatially-Random Discretization to pattern optimization (ablation) | SDCO · Digital world, Prob-E | ASR (%): 78.6 -> 100.0 | Improved |
| use SDCO instead of Gumbel-Softmax (ablation) | NORP optimization · Digital world, YOLOv11 T | ASR (%): 22.6 -> 99.4 | Improved |
| use SDCO instead of straight-through estimator (ablation) | NORP optimization · Digital world, YOLOv11 T | ASR (%): 86.8 -> 99.4 | Improved |
| use SDCO instead of straight-through estimator (ablation) | NORP optimization · Digital world, Prob-L | ASR (%): 94.4 -> 99.8 | Improved |
| use fusion-stage ensemble instead of single-detector training | NORP clothing · Transferability to Prob-L | ASR (%): 11.2 -> 99.4 | Improved |
| use fusion-stage ensemble instead of single-detector training | NORP clothing · Transferability to YOLOv11 | ASR (%): 1.0 -> 96.2 | Improved |
| use fusion-stage ensemble instead of single-detector training | NORP clothing · Transferability to AR-CNN | ASR (%): 71.2 -> 76.4 | Improved |
| use NORP adversarial clothing instead of UAP patch | adversarial attack · Physical world, Prob-E | ASR (%): 33.4 -> 73.5 | Improved |
| use NORP adversarial clothing instead of UAP patch | adversarial attack · Physical world, YOLOv11 RGB | ASR (%): 21.0 -> 61.2 | Improved |
| use NORP clothing instead of MAP | adversarial attack · Digital world, Prob-M | ASR (%): 37.2 -> 100.0 | Improved |
Closest papers by meaning
- Physical Adversarial Examples for Person Detectors in Thermal Images Based on 3D Modeling (TPAMI 2025 · similarity 0.90)
- Thermally Activated Dual-Modal Adversarial Clothing against AI Surveillance Systems (CVPR 2026 · similarity 0.89)
- Infrared Invisible Clothing: Hiding From Infrared Detectors at Multiple Angles in Real World (CVPR 2022 · similarity 0.89)
- Physically Realizable Natural-Looking Clothing Textures Evade Person Detectors via 3D Modeling (CVPR 2023 · similarity 0.88)
- Adversarial Attacks on Event-Based Pedestrian Detectors: A Physical Approach (AAAI 2025 · similarity 0.86)
- RAUCA: A Novel Physical Adversarial Attack on Vehicle Detectors via Robust and Accurate Camouflage Generation (ICML 2024 · similarity 0.84)
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.
