Research map / Adversarial robustness and security / Black-box and transfer attacks
GeoDA: A Geometric Framework for Black-Box Adversarial Attacks
Ali Rahmati, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard, Huaiyu Dai
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Restrict normal-vector queries to a low-frequency DCT subspace | GeoDA · ImageNet | median ℓ2 perturbation: approximately 27 % | Improved |
| Use an increasing, optimized query allocation across iterations instead of a single-iteration or uniform allocation (ablation) | GeoDA · ImageNet | ℓ2 perturbation | Improved |
| Move the sparse-attack starting point farther across the estimated boundary using ζ (ablation) | Sparse GeoDA · ImageNet | fooling rate and sparsity | Mixed |
| Use a larger query budget | Sparse GeoDA · ImageNet | sparse adversarial-example performance | Improved |
Closest papers by meaning
- CGBA: Curvature-aware Geometric Black-box Attack (ICCV 2023 · similarity 0.91)
- Exploiting the Local Parabolic Landscapes of Adversarial Losses to Accelerate Black-Box Adversarial Attack (ECCV 2022 · similarity 0.91)
- Adversarial Attacks on Black Box Video Classifiers: Leveraging the Power of Geometric Transformations (NeurIPS 2021 · similarity 0.90)
- Simple Black-box Adversarial Attacks (ICML 2019 · similarity 0.90)
- Black-box Adversarial Attacks with Limited Queries and Information (ICML 2018 · similarity 0.90)
- Guessing Smart: Biased Sampling for Efficient Black-Box Adversarial Attacks (ICCV 2019 · similarity 0.89)
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.
