通过梯度对齐提升伪造视频检测模型的泛化能力
RoGA: Towards Generalizable Deepfake Detection through Robust Gradient Alignment
- 用参数扰动使跨域梯度上升点对齐，增强模型鲁棒性
- 在多个数据集上超越当前最优方法，性能提升显著
- 无需额外正则项，适合实际部署的通用检测场景
深度伪造检测的领域泛化研究近年受到广泛关注，以往方法常引入额外模块防止对特定域的过拟合，但此类正则化可能干扰经验风险最小化（ERM）目标的优化，最终降低模型性能。本文提出一种新型学习目标，将泛化梯度更新与ERM梯度更新对齐。核心创新在于对模型参数施加扰动，使不同域的梯度上升点对齐，从而显著提升检测模型对领域迁移的鲁棒性。该方法有效保留域不变特征，同时管理域特定特征，且无需引入额外正则项。在多个具有挑战性的深度伪造检测数据集上的实验表明，本梯度对齐策略优于当前最先进的领域泛化技术，验证了方法的有效性。代码已开源。
原文摘要 · Abstract (English)
Recent advancements in domain generalization for deepfake detection have attracted significant attention, with previous methods often incorporating additional modules to prevent overfitting to domain-specific patterns. However, such regularization can hinder the optimization of the empirical risk minimization (ERM) objective, ultimately degrading model performance. In this paper, we propose a novel learning objective that aligns generalization gradient updates with ERM gradient updates. The key innovation is the application of perturbations to model parameters, aligning the ascending points across domains, which specifically enhances the robustness of deepfake detection models to domain shifts. This approach effectively preserves domain-invariant features while managing domain-specific characteristics, without introducing additional regularization. Experimental results on multiple challenging deepfake detection datasets demonstrate that our gradient alignment strategy outperforms state-of-the-art domain generalization techniques, confirming the efficacy of our method. The code is available at https://github.com/Lynn0925/RoGA.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。