Generating Attribution Reports for Manipulated Facial Images: A Dataset and Baseline
生成篡改面部图像的归因报告:一个数据集和基线
机构 * Xi’an Jiaotong University(西安交通大学) ; Hefei University of Technology(合肥工业大学) ; CSIRO(澳大利亚联邦科学与工业研究组织) ; Northwestern Polytechnical University(西北工业大学) ; University of Macau(澳门大学)
AI总结 本文提出多模态篡改追踪数据集和 ForgeryTalker 框架,通过联合定位篡改区域和生成自然语言解释,提升面部伪造检测的语义理解能力。
Comments Accepted to ACL 2026 (Main Conference). This version includes camera-ready revisions and updated experimental results
Journal ref Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026)