重新审视深度伪造检测：时间顺序持续学习与泛化的局限性
Revisiting Deepfake Detection: Chronological Continual Learning and the Limits of Generalization
- Sapienza University of Rome（罗马萨皮恩扎大学）
- University Ibn Khaldoun of Tiaret（蒂阿雷特伊本·khaldoun大学）
- University of Udine（乌迪内大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
该研究将深度伪造检测重构为持续学习问题，提出高效框架模拟7年时间演进，引入C-AUC和FWT-AUC指标，发现当前方法对未来生成器泛化接近随机，提出非通用深度伪造分布假设。
AI中文摘要：
深度伪造生成技术的快速演进对检测系统构成了严峻挑战，因为非持续学习方法需要频繁且昂贵的重新训练。我们将深度伪造检测（DFD）重新构建为一个持续学习（CL）问题，提出了一种高效框架，该框架能够增量适应新兴的视觉操纵技术，同时保留对过去生成器的知识。与先前依赖不真实模拟序列的方法不同，我们的框架在长达7年的扩展时间段内模拟了深度伪造技术在现实世界中的时间顺序演进。同时，我们的框架基于轻量级视觉骨干网络，以实现DFD系统的实时性能。此外，我们贡献了两个新指标：用于历史性能的持续AUC（C-AUC）和用于未来泛化的前向迁移AUC（FWT-AUC）。通过大量实验（超过600次模拟），我们实证表明，虽然高效适应（比完全重新训练快155倍）和对历史知识的稳健保留是可能的，但当前方法对未来生成器的泛化在没有额外训练的情况下仍接近随机（FWT-AUC ≈ 0.5），这是由于每个现有生成器特有的独特印记所致。这些观察是我们新提出的非通用深度伪造分布假设的基础。代码将在论文被接收后发布。
英文摘要：
The rapid evolution of deepfake generation technologies poses critical challenges for detection systems, as non-continual learning methods demand frequent and expensive retraining. We reframe deepfake detection (DFD) as a Continual Learning (CL) problem, proposing an efficient framework that incrementally adapts to emerging visual manipulation techniques while retaining knowledge of past generators. Our framework, unlike prior approaches that rely on unreal simulation sequences, simulates the real-world chronological evolution of deepfake technologies in extended periods across 7 years. Simultaneously, our framework builds upon lightweight visual backbones to allow for the real-time performance of DFD systems. Additionally, we contribute two novel metrics: Continual AUC (C-AUC) for historical performance and Forward Transfer AUC (FWT-AUC) for future generalization. Through extensive experimentation (over 600 simulations), we empirically demonstrate that while efficient adaptation (+155 times faster than full retraining) and robust retention of historical knowledge is possible, the generalization of current approaches to future generators without additional training remains near-random (FWT-AUC $\approx$ 0.5) due to the unique imprint characterizing each existing generator. Such observations are the foundation of our newly proposed Non-Universal Deepfake Distribution Hypothesis. \textbf{Code will be released upon acceptance.}
