🤖 AI Summary
This work addresses the dual challenge of accuracy and reasoning transparency in multimedia verification by proposing the first defeasible multi-agent framework that integrates multimodal large language models, external verification tools, and a quantified bipolar argumentation system (A-QBAF). The framework decomposes claims, retrieves supporting evidence, and constructs structured arguments annotated with source provenance and strength scores, enabling selective conflict resolution and uncertainty-aware belief revision within local argument graphs. By introducing A-QBAF into multimedia verification for the first time, the approach generates transparent, editable, and traceable verification reports that explicitly represent contested claims and their justifications. The system supports real-time user interaction and auditability, and has been released on an open-source platform to facilitate reproducibility and community adoption.
📝 Abstract
Multimedia verification requires not only accurate conclusions but also transparent and contestable reasoning. We propose a contestable multi-agent framework that integrates multimodal large language models, external verification tools, and arena-based quantitative bipolar argumentation (A-QBAF) as a submission to the ICMR 2026 Grand Challenge on Multimedia Verification. Our method decomposes each case into claim-centered sections, retrieves targeted evidence, and converts evidence into structured support and attack arguments with provenance and strength scores. These arguments are resolved through small local argument graphs with selective clash resolution and uncertainty-aware escalation. The resulting system generates section-wise verification reports that are transparent, editable, and computationally practical for real-world multimedia verification. Our implementation is public at: https://github.com/Analytics-Everywhere-Lab/MV2026_the_liems.