通过因果推理分离真实伪造特征，提升生成图像检测的泛化能力。
CausalCLIP: Causally-Informed Feature Disentanglement and Filtering for Generalizable Detection of Generated Images
- 构建因果模型，用掩码与独立性约束分离因果与非因果特征。
- 在未见过的生成模型上准确率提升6.83%，平均精度提升4.06%。
- 适合需要跨模型检测生成图像的研究者和安全应用开发者。
生成模型的快速发展带来了对跨多样、持续演进生成技术具备泛化能力的图像检测器的需求。然而，现有方法（包括利用预训练视觉语言模型的方法）常产生高度纠缠的表征，混合了任务相关的伪造线索（因果特征）与虚假或无关模式（非因果特征），从而限制了泛化性能。为此，我们提出CausalCLIP框架，显式地将因果特征与非因果特征解耦，并基于因果推断原则进行定向过滤，仅保留最具可迁移性和判别力的伪造线索。通过使用结构因果模型建模生成过程，并结合基于Gumbel-Softmax的特征掩码与希尔伯特-施密特独立性准则（HSIC）约束，强制特征间统计独立性，实现对鲁棒于分布偏移的稳定因果特征的隔离。在不同系列未见生成模型上的测试表明，CausalCLIP展现出强泛化能力，相比当前最优方法，准确率提升6.83%，平均精度提升4.06%。
原文摘要 · Abstract (English)
The rapid advancement of generative models has increased the demand for generated image detectors capable of generalizing across diverse and evolving generation techniques. However, existing methods, including those leveraging pre-trained vision-language models, often produce highly entangled representations, mixing task-relevant forensic cues (causal features) with spurious or irrelevant patterns (non-causal features), thus limiting generalization. To address this issue, we propose CausalCLIP, a framework that explicitly disentangles causal from non-causal features and employs targeted filtering guided by causal inference principles to retain only the most transferable and discriminative forensic cues. By modeling the generation process with a structural causal model and enforcing statistical independence through Gumbel-Softmax-based feature masking and Hilbert-Schmidt Independence Criterion (HSIC) constraints, CausalCLIP isolates stable causal features robust to distribution shifts. When tested on unseen generative models from different series, CausalCLIP demonstrates strong generalization ability, achieving improvements of 6.83% in accuracy and 4.06% in average precision over state-of-the-art methods.
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。