Research map / Adversarial robustness and security / Black-box and transfer attacks
LGV: Boosting Adversarial Example Transferability from Large Geometric Vicinity
Martin Gubri, Maxime Cordy, Mike Papadakis, Yves Le Traon, Koushik Sen
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 |
|---|---|---|---|
| replace single DNN surrogate with LGV (collect weights along high-lr SGD) | 1 DNN surrogate · ImageNet L-inf, target RN50 | success rate: 45.3 -> 95.4 | Improved |
| use SGD-collected LGV weights instead of random weight directions (RD) (ablation) | LGV · ImageNet L-inf, target RN50 | success rate: 60.6 -> 95.4 | Improved |
| use ensemble of 40 LGV models instead of their average (LGV-SWA) (ablation) | LGV · ImageNet L-inf, target RN50 | success rate: 84.9 -> 95.4 | Improved |
| add random directions around LGV-SWA (ablation) | LGV-SWA · ImageNet L-inf, 8 targets | success rate: +1.32 to +10.18 points (average 6.90) | Improved |
| add input diversity and momentum to LGV | LGV · ImageNet L-inf, target IncV3 | success rate: 25.3 -> 45.4 | Improved |
| add Ghost Networks to LGV | LGV · ImageNet L-inf, target RN50 | success rate: 95.4 -> 94.2 | Hurt |
| add SGM to LGV | LGV · ImageNet L-inf, target RN50 | success rate: 95.4 -> 86.9 | Hurt |
| add random Gaussian noise in weight space to single DNN (ablation) | 1 DNN · ImageNet L-inf, target RN50 | success rate: 45.3 -> 60.6 | Improved |
| shift LGV deviations to another solution (LGV-SWA + (LGV' - LGV-SWA')) (ablation) | LGV-SWA+RD · ImageNet L-inf | success rate: +3.27 to +12.32 points over LGV-SWA + RD (mean 8.61) | Improved |
| use unscaled instead of scaled (gamma 0.5) LGV deviations on regular DNN (ablation) | 1 DNN + gamma(LGV'-LGV-SWA') · ImageNet L-inf | success rate: -32.8 points on average | Hurt |
Closest papers by meaning
- LRS: Enhancing Adversarial Transferability through Lipschitz Regularized Surrogate (AAAI 2024 · similarity 0.90)
- Seeking Flat Minima Over Diverse Surrogates for Improved Adversarial Transferability: A Theoretical Framework and Algorithmic Instantiation (TPAMI 2026 · similarity 0.90)
- Boosting Adversarial Transferability through Augmentation in Hypothesis Space (CVPR 2025 · similarity 0.89)
- Enhancing Adversarial Transferability with Adversarial Weight Tuning (AAAI 2025 · similarity 0.88)
- Transferability Bound Theory: Exploring Relationship between Adversarial Transferability and Flatness (NeurIPS 2024 · similarity 0.88)
- Enhancing Adversarial Transferability with Checkpoints of a Single Model's Training (CVPR 2025 · similarity 0.88)
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.
