低秩正交子空间干预用于通用性面部伪造检测
Low-rank Orthogonal Subspace Intervention for Generalizable Face Forgery Detection
AI总结：
本文提出SeLop方法，通过低秩正交子空间干预消除虚假相关性，提升面部伪造检测的通用性和鲁棒性。
摘要：
通用性问题仍然是面部伪造检测中的关键挑战。本文探讨了Vanilla CLIP在"真实 vs. 假冒"检测中的通用性失败原因：在特征空间中少数主导的主成分主要编码了与伪造无关的信息，而非真实的伪造痕迹。然而，这种无关信息不可避免地导致了虚假相关性，严重限制了检测器的性能。我们将这种现象定义为"低秩虚假偏倚"。为了解决这个问题，我们从因果表示学习的角度提出了一个低秩表示空间干预范式，称为SeLop。SeLop将与伪造无关的虚假相关因素统一到一个低秩子空间中，并切断其与标签之间的统计捷径，从而将表示学习与真实的伪造痕迹对齐。具体来说，我们通过正交低秩投影将虚假相关特征分解为一个低秩子空间，然后从原始表示中移除该子空间，并训练其正交补集以捕捉与伪造相关的特征。这种低秩投影移除有效地消除了虚假相关因素，确保分类决策基于真实的伪造线索。仅需0.39M可训练参数，我们的方法在多个基准上实现了最先进的性能，展示了出色的鲁棒性和通用性。
英文摘要：
The generalization problem remains a key challenge in face forgery detection. This paper explores the reasons for the generalization failure of Vanilla CLIP: in ``real vs. fake" detection, the few dominant principal components in the feature space primarily encode forgery-irrelevant information, rather than authentic forgery traces. However, this irrelevant information inevitably leads to spurious correlations, severely limiting detector performance. We define this phenomenon as ``low-rank spurious bias". To address this, we propose a low-rank representation space intervention paradigm, named the SeLop, from the perspective of causal representation learning. SeLop unifies the spurious correlation factors irrelevant to forgery into a low-rank subspace and cuts off the statistical shortcut between it and the label, thus aligning representation learning with authentic forgery traces. Specifically, we decompose spurious correlation features into a low-rank subspace through orthogonal low-rank projection, then remove this subspace from the original representation and train its orthogonal complement to capture forgery-related features. This low-rank projection removal effectively eliminates spurious correlation factors, ensuring that classification decisions are based on authentic forgery cues. With only 0.39M trainable parameters, our method achieves state-of-the-art performance across several benchmarks, demonstrating excellent robustness and generalization.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
