Multi-Agent Forensic Reasoning for Generalizable Deepfake Video Detection

• 类型：arxiv
• 标识：2608.06865
• 链接：https://arxiv.org/abs/2608.06865
• 主分类：multimodal
• 形态：benchmark
• 被引：0
• 被引来源：Semantic Scholar
• S2被引：0
• 影响力被引：0
• TLDR：FaceVid-Forensics-100K is introduced, a large-scale deepfake video dataset comprising 100,000 videos and spanning 33 synthesis methods across face swapping, face reenactment, and entire-face synthesis, and a multi-agent forensic reasoning framework that employs four specialized domain-expert agents to independently analyze forgery cues from four perspectives.
• 副分类：agent
• 待LLM分类：否
• 标题中文：多智能体取证推理用于可泛化的深度伪造视频检测
• TLDR中文：提出 FaceVid-Forensics-100K，一个大规模深伪视频数据集，包含 100,000 个视频，涵盖 33 种合成方法，覆盖换脸、表情重演与全脸合成；同时提出一个多智能体取证推理框架，由四个领域专家 Agent 分别从四个角度独立分析伪造线索。
• 来源文件：
• /inbox/tom/_candidates/2026-08-11-rag-retrieval-reranking-candidates.json
• /inbox/tom/_candidates/2026-08-11-agent-rag-longcontext-candidates.json
• [S2 enrich]