Low-rank Orthogonal Subspace Intervention for Generalizable Face Forgery Detection

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wang, Chi, Hu, Xinjue, Wang, Boyu, He, Ziwen, Fu, Zhangjie
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912957169926144
author Wang, Chi
Hu, Xinjue
Wang, Boyu
He, Ziwen
Fu, Zhangjie
author_facet Wang, Chi
Hu, Xinjue
Wang, Boyu
He, Ziwen
Fu, Zhangjie
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2601_11915
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Low-rank Orthogonal Subspace Intervention for Generalizable Face Forgery Detection
Wang, Chi
Hu, Xinjue
Wang, Boyu
He, Ziwen
Fu, Zhangjie
Computer Vision and Pattern Recognition
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.
title Low-rank Orthogonal Subspace Intervention for Generalizable Face Forgery Detection
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.11915