Coordinated misleading text descriptions of video, audio, and meaning flip the appropriateness labels assigned by most multimodal LLMs in about 90% of test videos.
Adversaflow: Visual red team- ing for large language models with multi-level adversarial flow
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
baseline 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
baseline 1polarities
baseline 1representative citing papers
citing papers explorer
-
Watch, Listen, Understand, Mislead: Tri-modal Adversarial Attacks on Short Videos for Content Appropriateness Evaluation
Coordinated misleading text descriptions of video, audio, and meaning flip the appropriateness labels assigned by most multimodal LLMs in about 90% of test videos.