FMM-Attack: A Flow-based Multi-modal Adversarial Attack on Video-based LLMs.

Jinmin Li Kuofeng Gao Yang Bai Jingyun Zhang Shutao Xia Yisen Wang

Published in: CoRR (2024)

Keyphrases

multi modal
multi modality
image annotation
cross modal
audio visual
semantic concepts
video search
humanoid robot
metadata
image segmentation
relevance feedback
mr images
uni modal