计算机科学与探索2026，Vol.20Issue(6)：1702-1715,14.DOI:10.3778/j.issn.1673-9418.2504051
基于预训练的人脸图像伪造算法识别模型
Recognition Model of Face Image Forgery Algorithm Based on Pre-training
摘要
Abstract
With the development of deep learning,face forgery methods are gradually increasing.The existing deepfake face detection models are mostly concentrated on the true or false binary classification,with poor generalization,which is difficult to adapt to the current iterative forgery algorithms.What's more,the researches on the face images with different sharpness and noise interferences are not deep enough.Therefore,in order to solve the above problems,this paper proposes a muti-class face image forging algorithm recognition method based on pre-training model.Specifically,the entire model is divided into two major modules:pre-training and main training.In the pre-training stage,perception and reconstruction-masked autoencoders(PR-MAE)based on mask strategy is designed to further excavate mask features to construct percep-tual loss and enhance the global perceptual learning ability of the model.And the UDenseNet training network is con-structed to improve the ability to capture details by U-shaped dense blocks,in order to adapt to more complex tasks.Mean-while,the multi-noise fusion(MNF)module is added to improve the model's resistance to multiple types of noise by dyna-mic injection strategy and enhance the robustness.In the main training process,CLIP supplementary learning module is introduced to perform the comparative learning with the pre-trained UDenseNet to improve the accuracy of the model.Experimental results show that the accuracy(ACC)index of this method can reach 89.12%and 90.25%on the latest MCFF and DF40 classification datasets,the average ACC index can reach 87.22%in the tests of the low definition images,83.36%in the tests of various kinds of noise and the index can reach more than 90%at the highest in the generalization test.关键词
伪造算法识别/人脸伪造/深度学习Key words
forgery algorithm recognition/face forgery/deep learning分类
信息技术与安全科学引用本文复制引用
丁博文,芦天亮,彭舒凡,王珑皓..基于预训练的人脸图像伪造算法识别模型[J].计算机科学与探索,2026,20(6):1702-1715,14.基金项目
公安部科技计划项目(2023JSM09) （2023JSM09）
公安行业大模型研究与应用实验室技术研发项目(2024300050036). This work was supported by the Science and Technology Project of Ministry of Public Security of China(2023JSM09),and the Techni-cal Research and Development Project of the Research and Application Laboratory for Large Models in the Public Security Industry(2024300050036). （2024300050036）
