arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.12042eess.AS
收藏

FakeMark:基于水印伪影的深度伪造语音溯源

FakeMark: Deepfake Speech Attribution With Watermarked Artifacts

National Institute of Informatics(信息学研究所)

Wanying Ge, Xin Wang, Junichi Yamagishi

AI总结:

FakeMark提出一种注入伪影相关水印的深度伪造语音溯源框架,结合水印与内在伪影提升跨域泛化及抗失真能力。

摘要:

深度伪造语音溯源对现有解决方案而言仍具挑战性。基于分类器的方法往往难以泛化到域偏移样本,而基于水印的方法则容易被编解码压缩或恶意移除攻击等失真所破坏。为解决这些问题,我们提出FakeMark,一种新颖的水印框架,其注入与深度伪造系统相关联的伪影相关水印,而非预先指派的比特串消息。该设计使检测器能够利用注入的水印和内在的深度伪造伪影来溯源源系统,即使其中一种线索缺失或被移除,仍能保持有效。实验结果表明,FakeMark在基于分类器的方法难以处理的跨数据集样本上提升了泛化能力,并在基于传统水印方法失效的各种失真条件下保持了高准确率。

英文摘要:

Deepfake speech attribution remains challenging for existing solutions. Classifier-based solutions often fail to generalize to domain-shifted samples, and watermarking-based solutions are easily compromised by distortions like codec compression or malicious removal attacks. To address these issues, we propose FakeMark, a novel watermarking framework that injects artifact-correlated watermarks associated with deepfake systems rather than pre-assigned bitstring messages. This design allows a detector to attribute the source system by leveraging both injected watermark and intrinsic deepfake artifacts, remaining effective even if one of these cues is elusive or removed. Experimental results show that FakeMark improves generalization to cross-dataset samples where classifier-based solutions struggle and maintains high accuracy under various distortions where conventional watermarking-based solutions fail.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑