Research map / Adversarial robustness and security / Black-box and transfer attacks
Adversarial Attacks on Black Box Video Classifiers: Leveraging the Power of Geometric Transformations
Shasha Li, Abhishek Aich, Shitong Zhu, Salman Asif, Chengyu Song, Amit Roy-Chowdhury, Srikanth Krishnamurthy
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 |
|---|---|---|---|
| GeoTrap instead of Motion Sampler | GeoTrap · Jester untargeted, SlowFast | queries: 1906 -> 521 | Improved |
| GeoTrap instead of Heuristic attack | GeoTrap · untargeted attacks, 4 victims | success rate: about 6% higher | Improved |
| GeoTrap instead of Motion Sampler | GeoTrap · untargeted attacks | average number of queries: 45% fewer | Improved |
| GeoTrap instead of Heuristic attack | GeoTrap · targeted attacks | success rate: over 30% higher | Improved |
| GeoTrap instead of Motion Sampler | GeoTrap · targeted attacks | success rate: over 8% higher | Improved |
| GeoTrap instead of Heuristic attack | GeoTrap · Jester targeted, TPN | success rate: 44.4 -> 92.6 | Improved |
| Heuristic attack instead of GeoTrap | GeoTrap · Jester targeted, TPN | queries | Hurt |
| Dilation transformation instead of translation-dilation (ablation) | GeoTrap · Jester untargeted, SlowFast | queries: 521 -> 325 | Improved |
| Affine transformation instead of translation-dilation (ablation) | GeoTrap · Jester untargeted, SlowFast | queries: 521 -> 1057 | Hurt |
| Similarity transformation instead of translation-dilation (ablation) | GeoTrap · Jester untargeted, C3D | queries: 1602 -> 1621 | No clear effect |
| Translation only instead of translation-dilation (ablation) | GeoTrap · Jester untargeted, C3D | queries: 1602 -> 3340 | Hurt |
| Affine instead of translation-dilation (ablation) | GeoTrap · UCF-101 untargeted, C3D | success rate: 86.2 -> 84.8 | Hurt |
Closest papers by meaning
- GeoDA: A Geometric Framework for Black-Box Adversarial Attacks (CVPR 2020 · similarity 0.90)
- Heuristic Black-Box Adversarial Attacks on Video Recognition Models (AAAI 2020 · similarity 0.89)
- Exploiting the Local Parabolic Landscapes of Adversarial Losses to Accelerate Black-Box Adversarial Attack (ECCV 2022 · similarity 0.88)
- Guessing Smart: Biased Sampling for Efficient Black-Box Adversarial Attacks (ICCV 2019 · similarity 0.87)
- A General Framework for Black-Box Attacks Under Cost Asymmetry (ICLR 2026 · similarity 0.87)
- Simple Black-box Adversarial Attacks (ICML 2019 · similarity 0.87)
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.
