📄 中文摘要
随着高度逼真的面部伪造技术的普及，迫切需要有效的检测方法。然而，现有方法往往由于生成样本的伪造技术多样性导致显著的分布偏移，从而在准确性和泛化能力上存在不足。为了解决这些挑战，提出了一种新颖的多变量软混合增强与CLIP引导的伪造强度估计框架（MSBA-CLIP）。该方法利用CLIP的多模态对齐能力捕捉微妙的伪造痕迹。引入的多变量软混合增强（MSBA）策略通过随机权重混合来自多种方法的伪造图像，迫使模型学习可泛化的模式。
基于多变量软混合和CLIP图像-文本对齐的深伪检测
📄 English Summary
Detecting Deepfakes with Multivariate Soft Blending and CLIP-based Image-Text Alignment
The proliferation of highly realistic facial forgeries necessitates robust detection methods. Existing approaches often suffer from limited accuracy and poor generalization due to significant distribution shifts among samples generated by diverse forgery techniques. To address these challenges, a novel Multivariate and Soft Blending Augmentation with CLIP-guided Forgery Intensity Estimation (MSBA-CLIP) framework is proposed. This method leverages the multimodal alignment capabilities of CLIP to capture subtle forgery traces. The introduced Multivariate and Soft Blending Augmentation (MSBA) strategy synthesizes images by blending forgeries from multiple methods with random weights, compelling the model to learn generalizable patterns.