Recent advancements in face forgery techniques have continuously evolved, leading to emergent security concerns in society. Existing detection methods have poor generalization ability due to the insufficient extraction of dynamic inconsistency cues on the one hand, and their inability to deal well with the gaps between forgery techniques on the other hand. To develop a new generalized framework that emphasizes extracting generalizable multi-timescale inconsistency cues. Firstly, we capture subtle dynamic inconsistency via magnifying the multipath dynamic inconsistency from the local-consecutive short-term temporal view. Secondly, the inter-group graph learning is conducted to establish the sufficient-interactive long-term temporal view for capturing dynamic inconsistency comprehensively. Finally, we design the domain alignment module to directly reduce the distribution gaps via simultaneously disarranging inter- and intra-domain feature distributions for obtaining a more generalized framework. Extensive experiments on six large-scale datasets and the designed generalization evaluation protocols show that our framework outperforms state-of-the-art deepfake video detection methods.

人脸伪造技术的最新进展不断发展，导致社会上出现了安全问题。现有的检测方法一方面由于动态不一致线索提取不足，另一方面无法很好地处理伪造技术之间的差距，泛化能力较差。开发一个新的通用框架，强调提取可推广的多时间尺度不一致线索。首先，我们通过从局部连续的短期时间视图中放大多径动态不一致性来捕捉细微的动态不一致。其次，进行组间图学习，建立充分交互的长期时间视图，以全面捕获动态不一致。最后，我们设计了域对齐模块，通过同时解调域间和域内特征分布来直接减少分布差距，以获得更通用的框架。对 6 个大规模数据集和设计的泛化评估协议的广泛实验表明，我们的框架优于最先进的 deepfake 视频检测方法。