多智能體取證推理於可泛化深度偽造影片偵測之應用

摘要

惡意利用生成式人工智慧製造高度逼真的深度偽造影片，引發嚴重的倫理疑慮，並對人工智慧安全構成重大挑戰。然而，現有的深度偽造影片基準測試對近期合成方法的涵蓋範圍有限，且普遍缺乏可靠的細粒度文字註解。與此同時，傳統偵測器與多模態大型語言模型（MLLMs），無論是單一模型運作或依賴單一分析視角，往往無法捕捉細微的偽造痕跡，因而限制了其對新興人工智慧生成方法的泛化能力。為了解決這些限制，我們提出了 FaceVid-Forensics-100K，這是一個大規模的深度偽造影片資料集，包含 100,000 部影片，涵蓋 33 種合成方法，橫跨人臉交換、人臉重演與全臉合成，包括 Seedance 2.0 等近期生成器。該資料集提供了視覺觀察的細粒度文字註解，以及與判決一致的鑑識解釋；這些內容是透過一個由先進 MLLMs 驅動的多模型彙整與衝突解決流程自動合成而成。在此基準的基礎上，我們提出一個多代理鑑識推理框架，採用四個專門的領域專家代理，從紋理、光線、動作與物理四個面向獨立分析偽造線索。接著，一個裁判代理會整合這些報告，產出最終預測與解釋。在域外測試集上的廣泛評估顯示，儘管我們框架完全由小型開源 MLLMs 組成，其效能仍優於所有方法，包括閉源 GPT 與 Gemini 模型，並且在此基準的所有回報指標上名列第一。專案頁面位於 https://xavierjiezou.github.io/ARGUS/。

English

The malicious use of generative artificial intelligence to create highly realistic deepfake videos raises serious ethical concerns and poses substantial challenges to AI safety. However, existing deepfake video benchmarks provide limited coverage of recent synthesis methods and generally lack reliable fine-grained textual annotations. Meanwhile, conventional detectors and multimodal large language models (MLLMs), whether operating as a single model or relying on a single analytical perspective, often fail to capture subtle forgery artifacts, limiting their generalization to emerging AI-generated methods. To address these limitations, we introduce FaceVid-Forensics-100K, a large-scale deepfake video dataset comprising 100,000 videos and spanning 33 synthesis methods across face swapping, face reenactment, and entire-face synthesis, including recent generators such as Seedance 2.0. The dataset provides fine-grained textual annotations of visual observations and verdict-consistent forensic explanations, automatically synthesized through a multi-model aggregation and conflict-resolution pipeline powered by advanced MLLMs. Building on this benchmark, we propose a multi-agent forensic reasoning framework that employs four specialized domain-expert agents to independently analyze forgery cues from four perspectives: texture, lighting, motion, and physics. A judge agent then reconciles their reports to produce a final prediction together with an explanation. Extensive evaluations on out-of-domain test sets show that, despite being composed entirely of small open-source MLLMs, our framework outperforms all methods including closed-source GPT and Gemini models and ranks first across all reported metrics on this benchmark. The project page is available at https://xavierjiezou.github.io/ARGUS/.