Towards Generalizable Deepfake Detection by Primary Region Regularization
Open Access
- 9 February 2026
- journal article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Multimedia Computing, Communications, and Applications
- Vol. 22 (2) , 1-25
Abstract
The existing deepfake detection methods have reached a bottleneck in generalizing to unseen forgeries and manipulation approaches. Based on the observation that the deepfake detectors exhibit a preference for overfitting specific primary regions in input, this article enhances the generalization capability from a novel regularization perspective. This can be simply achieved by augmenting the images through primary region removal, thereby preventing the detector from over-relying on data bias. Our method consists of two stages, namely the static localization for primary region maps, as well as the dynamic exploitation of primary region masks. The proposed method can be seamlessly integrated into different backbones without affecting their inference efficiency. We conduct extensive experiments over five widely used deepfake datasets—DFDC, DF-1.0, Celeb-DF, WildDF, and FFIW with seven backbones. Our method demonstrates an average performance improvement of 6% across different backbones and performs competitively with several state-of-the-art baselines.Keywords
All Related Versions
Funding Information
- Ministry of Education, Singapore (MOE-MOET32022-0001)
This publication has 47 references indexed in Scilit:
- Face Reconstruction-Based Generalized Deepfake Detection Model with Residual Outlook AttentionACM Transactions on Multimedia Computing, Communications, and Applications, 2025
- Deepfake Detection Using Spatiotemporal TransformerACM Transactions on Multimedia Computing, Communications, and Applications, 2024
- Detecting Deepfake Videos using Spatiotemporal Trident NetworkACM Transactions on Multimedia Computing, Communications, and Applications, 2024
- Head Pose Estimation Patterns as Deepfake DetectorsACM Transactions on Multimedia Computing, Communications, and Applications, 2024
- Continual Face Forgery Detection via Historical Distribution PreservingInternational Journal of Computer Vision, 2024
- Supervised Anomaly Detection via Conditional Generative Adversarial Network and Ensemble Active LearningIEEE Transactions on Pattern Analysis and Machine Intelligence, 2022
- FakeLocator: Robust Localization of GAN-Based Face ManipulationsIEEE Transactions on Information Forensics and Security, 2022
- What Makes Fake Images Detectable? Understanding Properties that Generalize2020
- A survey of regularization strategies for deep modelsArtificial Intelligence Review, 2019
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015