DVAR: Adversarial Multi-Agent Debate for Video Authenticity Detection

2026-04-18

AI Summary

This work addresses the limited generalization of existing video authenticity detection methods when confronted with out-of-distribution, novel generative models. To overcome this challenge, the authors propose DVAR, a novel training-free multi-agent adversarial debate framework. DVAR enables two agents—one advocating a “generative hypothesis” and the other defending a “natural mechanism”—to engage in cross-examination and logical reasoning to assess video authenticity. The framework incorporates the Minimum Description Length (MDL) principle to quantify explanatory costs and integrates a dynamic knowledge base, GenVideoKB, to enhance reasoning capabilities. Evaluated on unseen generative models, DVAR significantly outperforms current supervised approaches while delivering transparent, interpretable, and traceable reasoning pathways.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Web data provenance, reliability, and authenticity
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

The rapid evolution of video generation technologies poses a significant challenge to media forensics, as conventional detection methods often fail to generalize beyond their training distributions. To address this, we propose DVAR (Debate-based Video Authenticity Reasoning), a training-free framework that reformulates video detection as a structured multi-agent forensic reasoning process. Moving beyond the paradigm of pattern matching, DVAR orchestrates a competition between a Generative Hypothesis Agent and a Natural Mechanism Agent. Through iterative rounds of cross-examination, these agents defend their respective explanations against abnormal evidence, driving a logical convergence where the truth emerges from rigorous stress-testing. To adjudicate these conflicting claims, we apply Occam's Razor through the Minimum Description Length (MDL) framework, defining an Explanatory Cost to quantify the "logical burden" of each reasoning path. Furthermore, we integrate GenVideoKB, a dynamic knowledge repository that provides high-level reasoning heuristics on generative boundaries and failure modes. Extensive experiments demonstrate that DVAR achieves competitive performance against supervised state-of-the-art methods while exhibiting superior generalization to unseen generative architectures. By transforming detection into a transparent debate, DVAR provides explicit, interpretable reasoning traces for robust video authenticity assessment.

Problem

• video authenticity detection
• generalization
• generative video
• media forensics
• unseen generative architectures

Innovation

• adversarial multi-agent debate
• training-free detection
• Minimum Description Length
• explainable reasoning
• video authenticity