AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Motion-Excited Sampler: Video Adversarial Attack with Sparked Prior

Hu Zhang, Linchao Zhu, Yi Zhu, Yi Yang

ECCV 2020 · Adversarial robustness and security › Attacks on text and graphs · approach: Adversarial perturbations beyond images · 34 citations (OpenAlex)

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 changedWhereMeasuredEffect
add motion-excited sampler (motion maps) instead of per-frame random noise (ablation)ME-Sampler · UCF-101 vs I3DANQ: 15,798 -> 6,876Improved
add motion maps instead of one noise replicated to all frames (ablation)ME-Sampler · UCF-101 vs I3DANQ: 22,908 -> 6,876Improved
add motion maps instead of one noise replicated to all frames (ablation)ME-Sampler · UCF-101 vs I3Dsuccess rate: 93.33 -> 100.0Improved
use true motion map instead of uniformly sampled values with same binary mask (ablation)ME-Sampler · UCF-101 vs TSN2DANQ and success rate: 9,319 queries saved and 16.77% higher success rateImproved
use sequenced-value map instead of uniform-random map (ablation)ME-Sampler · Kinetics-400 vs I3DANQ and success rate: saves more than 1,000 queries but gets 2% higher success rateImproved
use true motion map instead of sequenced-value map (ablation)ME-Sampler · UCF-101 vs I3DANQ: 11,471 -> 6,876Improved
use original motion vector instead of randomized motion map (ablation)ME-Sampler · UCF-101 vs I3DANQ: 6,970 -> 6,876No clear effect
use motion vector instead of optical flowME-Sampler · UCF-101 vs TSN2DANQ: 6,598 -> 5,353Improved
use motion vector instead of optical flowME-Sampler · HMDB-51 vs I3DANQ: 3,306 -> 3,915Hurt
use ME-Sampler (MV) instead of V-BADME-Sampler · SthSth-V2 vs I3DANQ and success rate: 7,239 -> 592 queries, 97.70 -> 100.0 SRImproved
use ME-Sampler (MV) instead of E-BanditsME-Sampler · UCF-101 vs TSN2DANQ and success rate: 24,890 -> 5,353 queries, 66.33 -> 99.00 SRImproved
use ME-Sampler instead of V-BAD in targeted attackME-Sampler · HMDB-51 targeted vs TSN2Dsuccess rate: 64.86 -> 92.16Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy