Adding LVLM-generated captions, retrieved evidence, and emotion-specific expert processors to a multimodal news classifier improves fake news detection accuracy on Weibo and Twitter benchmarks.
Cross-modal ambiguity learning for multimodal fake news detection
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.MM 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
KEN: Knowledge Augmentation and Emotion Guidance Network for Multimodal Fake News Detection
Adding LVLM-generated captions, retrieved evidence, and emotion-specific expert processors to a multimodal news classifier improves fake news detection accuracy on Weibo and Twitter benchmarks.