MSU GenText-Forensics 挑战赛 2026 技术报告

中文摘要 AI 辅助

文档文本伪造已超越简单的像素级操作：现代攻击不仅改变文档的外观，还改变其含义，并日益针对消费此类文档的 OCR 与 LLM 流水线。因此，ACM MM 2026 GenText-Forensics 挑战赛要求系统不仅判断多语言文本图像是否被伪造，还要定位篡改点、识别攻击类型，并生成带有支持证据的可读取证报告。我们提出了一种分解式思维链（CoT）流水线，结合文档篡改检测器（DTD）与两个 Qwen3-VL-32B 视觉语言模型，每个模型通过 LoRA 适配到不同的子任务。DTD 生成篡改概率图，并转换为编号的候选区域；第一个模型（过滤器）验证这些区域并分配初步的伪造类型，而第二个模型（语义侦探）合并并重新定位存活的区域，搜索像素级检测器不可见的纯语义异常，并撰写最终报告。两个模型均通过从特权 Qwen3-VL-235B 教师模型（可访问真实掩码和报告）中蒸馏思维链轨迹进行训练。我们的方法在 ACM MM 2026 GenText-Forensics 挑战赛中获得第三名。我们详细描述了数据准备、测试时增强、区域渲染、蒸馏协议和训练配置，并报告了检测器阈值、提示设计和流水线分解的消融实验。

英文摘要

Document text forgery has evolved beyond simple pixel-level manipulation: modern attacks alter not only the appearance of a document but also its meaning, and increasingly target the OCR & LLM pipelines that consume such documents. The ACM MM 2026 GenText-Forensics challenge therefore requires systems that not only decide whether a multilingual text image is forged, but also localize the point of manipulation, identify the attack type, and produce a human-readable forensic report with supporting evidence. We present our solution, a decomposed chain-of-thought (CoT) pipeline that combines a document tampering detector (DTD) with two Qwen3-VL-32B vision-language models, each LoRA-adapted to a distinct sub-task. DTD produces tampering probability maps that are converted into numbered candidate regions; a first model (the Filterer) validates these regions and assigns a preliminary forgery type, while a second model (the Semantic Detective) merges and re-grounds the surviving regions, searches for purely semantic anomalies that are invisible to pixel-level detectors, and writes the final report. Both models are trained by distilling chain-of-thought traces from a privileged Qwen3-VL-235B teacher that has access to ground-truth masks and reports. Our approach secured third place in the ACM MM 2026 GenText-Forensics challenge. We describe the data preparation, test-time augmentation, region rendering, distillation protocol, and training configuration in detail, and report ablations over detector thresholds, prompt designs, and pipeline decompositions.

发表机构

• Lomonosov Moscow State University（莫斯科国立大学）
• MSU Institute for Artificial Intelligence（莫斯科国立大学人工智能研究所）

机构由 AI 辅助整理，请以论文原文为准。