EUVIMEDEUVIMED entdecken
Crossref

IN 3 SIGHT: Toward Cognitive Forensic Reasoning for OOC Misinformation Detection

Kaige Li, Zhaowei Wu, Chao Huang, Wei Wang, Xiaochun Cao

IEEE Transactions on Image Processing · 2026

DOI: 10.1109/tip.2026.3736797 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Image Processing, 2026. Verfasst von Kaige Li und 4 weiteren Autor:innen. Hauptthemen laut MEDLINE: Image Processing, Computer-Assisted, Forensic Sciences, Communication. Die Arbeit nennt 55 Literaturangaben. Gefördert von: National Natural Science Foundation of China, China Postdoctoral Science Foundation General Fund, Shenzhen Key Laboratory of Media Security.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Out-of-context (OOC) misinformation, where genuine images are paired with misleading text, poses substantial societal risks due to its deceptive narratives. Prior efforts primarily assess image-text consistency but often lack explainable judgments, limiting their usefulness for forensic validation. Recently, Multimodal Large Language Models (MLLMs) show promise with vast world knowledge and strong reasoning capabilities, yet their judgments in OOC detection remain unreliable. On one hand, their static internal knowledge quickly becomes outdated in fast-moving news domains, and compensating with externally retrieved evidence often introduces noise and conflicts with their internal priors. On the other hand, the absence of task-specific cognitive calibration leads to hallucinations, whereas instruction tuning introduces such specialization only at the cost of vast curated data, rendering it impractical for dynamic, real-world scenarios. In this paper, we propose $\text {IN}^{3}$ SIGHT, a cognitive framework that emulates expert forensic reasoning through three synergistic stages: Epistemic Inspection, which calibrates the epistemic state of an MLLM by task-aware prompt optimization and discrete confidence stratification, yielding a stable epistemic anchor without model tuning; Semantic Investigation, which audits multimodal evidence by enforcing event-level identity and deriving a relevance-aware external assessment under conditional visual anchoring, distilling retrieved evidence into claim-aligned snippets and summaries; and Dialectical Introspection, which conducts multi-fact reasoning by dynamically adjusting the scope of consulted evidence according to evidential certainty and produces the final consistency verdict. Experiments show that $\text {IN}^{3}$ SIGHT markedly improves zero-shot MLLM performance by 15.0 percentage points without parameter updates and achieves competitiveness with or even surpasses instruction-tuned methods. This indicates that imposing structured cognitive constraints and evidence-grounded reasoning on MLLMs offers a principled pathway toward reliable misinformation forensics.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Image Processing, Computer-AssistedForensic SciencesCommunicationLarge Language ModelsHumansAlgorithms

Förderung

  • National Natural Science Foundation of China
  • China Postdoctoral Science Foundation General Fund
  • Shenzhen Key Laboratory of Media Security
  • Open Research Fund of State Key Laboratory of Internet Architecture
  • Shenzhen Science and Technology Program

Zitieren

APA

Li, K., Wu, Z., Huang, C., Wang, W., & Cao, X. (2026). IN 3 SIGHT: Toward Cognitive Forensic Reasoning for OOC Misinformation Detection. IEEE Transactions on Image Processing. https://doi.org/10.1109/tip.2026.3736797

Vancouver

Li K, Wu Z, Huang C, Wang W, Cao X. IN 3 SIGHT: Toward Cognitive Forensic Reasoning for OOC Misinformation Detection. IEEE Transactions on Image Processing. 2026. doi:10.1109/tip.2026.3736797

BibTeX
@article{Li2026,
  author = {Li, Kaige and Wu, Zhaowei and Huang, Chao and Wang, Wei and Cao, Xiaochun},
  title = {IN 3 SIGHT: Toward Cognitive Forensic Reasoning for OOC Misinformation Detection},
  journal = {IEEE Transactions on Image Processing},
  year = {2026},
  doi = {10.1109/tip.2026.3736797}
}
Kostenlos starten →