Research map / Adversarial robustness and security / Black-box and transfer attacks
Boosting the Transferability of Video Adversarial Examples via Temporal Translation
Read the paperCheck 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 |
|---|---|---|---|
| Add temporal translation to TI(1) | TI(1) · UCF-101 | average ASR (%): 45.96 -> 52.48 | Helped |
| Add temporal translation to TI(1) | TI(1) · Kinetics-400 | average ASR (%): 54.38 -> 63.22 | Helped |
| Add temporal translation to ATA(10) | ATA(10) · UCF-101 | average ASR (%): 9.98 -> 12.87 | Helped |
| Add temporal translation to ATA(10) | ATA(10) · Kinetics-400 | average ASR (%): 8.52 -> 10.58 | Helped |
| Add temporal translation to MI(10) | MI(10) · UCF-101 | average ASR (%): 33.58 -> 44.80 | Helped |
| Add temporal translation to MI(10) | MI(10) · Kinetics-400 | average ASR (%): 40.60 -> 63.50 | Helped |
| Use linear instead of uniform gradient weights (ablation) | TT(1), NL-101 white-box · TT(1), NL-101 white-box | average ASR (%): 65.06 -> 66.37 | Helped |
| Use Gaussian instead of uniform gradient weights (ablation) | TT(1), NL-101 white-box · TT(1), NL-101 white-box | average ASR (%): 65.06 -> 66.80 | Helped |
| Use linear instead of uniform gradient weights (ablation) | TT(10), NL-101 white-box · TT(10), NL-101 white-box | average ASR (%): 49.80 -> 52.60 | Helped |
| Use adjacent rather than random or remote shifts (ablation) | TT, NL-101 white-box · TT(1), NL-101 white-box | average ASR (%): 58.06 -> 65.56; 57.31 -> 65.56 | Helped |
| Use adjacent rather than random or remote shifts (ablation) | TT, NL-101 white-box · TT(10), NL-101 white-box | average ASR (%): 23.25 -> 43.37; 26.81 -> 43.37 | Helped |
| Increase temporal shift length L (ablation) | TT(1) and TT(10) · Black-box models other than NL-50 | ASR | Helped |
Closest papers by meaning
- Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks (CVPR 2019 · similarity 0.91)
- Delving into Transferable Adversarial Examples and Black-box Attacks (ICLR 2017 · similarity 0.91)
- Transferability of Adversarial Attacks in Video-based MLLMs: A Cross-modal Image-to-Video Approach (AAAI 2026 · similarity 0.90)
- Boosting Adversarial Transferability using Dynamic Cues (ICLR 2023 · similarity 0.90)
- Rethinking Model Ensemble in Transfer-based Adversarial Attacks (ICLR 2024 · similarity 0.88)
- Boosting Adversarial Transferability through Augmentation in Hypothesis Space (CVPR 2025 · 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.