Multimodal fake news detection requires both finegrained text-image alignment and external factual context. This paper presents a three-stage framework centered on modality-specific global-token cross-attention. The proposed attention mechanism independently parameterizes text-to-image and image-to-text projections to...
LCFND, a staged and fully specified bounded-evidence verification pipeline for localized conflict discovery, structured evidence generation, and graph-based verification, is presented, better suited to offline or high-risk verification than to large-scale real-time screening.
Yuechuan Zhang, Hongyu Jin, Ya-Xuan Wang et al.· Journal of King Saud Univers...· 0 citations
TERN induces latent deception types from image-side multimodal features through prototype-based clustering, uses the induced assignments as structural priors for downstream veracity prediction, disentangles type-discriminative factors from semantic content, and performs type-conditioned hierarchical reasoning over text...
