A multi-agent RL black-box attack that localizes sensitive frames and patches in videos achieves lower L1 perturbation and query counts than prior video attacks on HMDB-51 and UCF-101.
Appending adversarial frames for universal video at - tack
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Robustness Evaluation for Video Models with Reinforcement Learning
A multi-agent RL black-box attack that localizes sensitive frames and patches in videos achieves lower L1 perturbation and query counts than prior video attacks on HMDB-51 and UCF-101.