基于域对抗学习的可泛化虚假人脸检测方法研究

翁泽佳; 陈静静; 姜育刚

doi:10.7544/issn1000-1239.2021.20200803

基于域对抗学习的可泛化虚假人脸检测方法研究

(复旦大学计算机科学技术学院上海 201203) (上海市智能信息处理重点实验室(复旦大学计算机科学技术学院) 上海 200433) (zjweng20@fudan.edu.cn)

基金项目: 国家自然科学基金项目(62032006)

详细信息

中图分类号: TP391
计量
- 文章访问数: 683
- HTML全文浏览量: 10
- PDF下载量: 341
出版历程
- 发布日期: 2021-06-30

On the Generalization of Face Forgery Detection with Domain Adversarial Learning

(School of Computer Science, Fudan University, Shanghai 201203) (Shanghai Key Laboratory of Intelligent Information Processing (School of Computer Science, Fudan University), Shanghai 200433)

Funds: This work was supported by the National Natural Science Foundation of China (62032006).

摘要

摘要: 随着生成式对抗网络(generative adversarial networks, GAN)的快速发展，虚假人脸生成技术取得了显著进展.为了降低以假乱真的人脸生成技术给社会带来的危害，虚假人脸鉴别成为一个非常重要的课题，吸引了国内外研究者的广泛关注.然而，目前虚假人脸鉴别的研究工作相对较少，仍然有许多问题需要被解决.其中如何提升鉴别模型的迁移泛化能力是至关重要的问题，也是虚假人脸检测任务能否实际投入使用的关键所在.如何提升虚假人脸鉴别方法的泛化能力，即做到在没有见过的生成方法产生的数据上仍然准确有效非常重要.对此，提出了基于域对抗学习的可泛化虚假人脸检测模型，通过引入领域对抗分支，弱化特征提取器对于特定生成模型非鲁棒性特征的提取，模型能够抽取鲁棒性更强、泛化能力更高的特征，从而在没有见过的生成方法产生的虚假人脸图片上具有更好的鉴别表现.实验结果表明：所提出的方法能够提升鉴别模型的泛化能力，显著提升虚假人脸鉴别模型在未知生成模型产生的虚假图像上的性能.
- 虚假人脸检测 /
- 域自适应 /
- 域对抗学习 /
- 鲁棒特征学习 /
- 泛化性
Abstract: With the rapid development of generative adversarial networks (GAN), breakthrough progress has been made in fake face generation. In order to reduce the harmful effects of fake face generation technology to society, fake face identification has become a very important topic, which has attracted numerous research efforts. Although impressive progress has been made in fake face identification, there are still many problems to be solved. Among them, how to improve the generalization ability of the fake face detection model is a crucial issue, and it is also the key to deploy fake face detection techniques in real-world scenarios. This paper studies the fake face identification problem, aiming to improve the generalization ability of the fake face identification model. Motivated by the idea of unsupervised domain adaptation, this paper introduces the domain adversarial branch to weaken the extraction of non-robust features of specific generative models by the feature extractor, so that the model can extract features with stronger robustness and higher generalization ability, improving the identification performance on the fake face images generated by unknown GANs. Experimental results show that the method proposed in this paper can effectively improve the generalization ability of the identification model, and significantly improve the performance of the fake face identification model on the fake images generated by the unknown generation model.
- face forgery detection /
- domain adaptation /
- domain adversarial learning /
- robust feature learning /
- generalization

HTML全文

参考文献(0)

施引文献(13)

期刊类型引用(7)

1.	李志博，李清宝，兰明敬. 基于ART优化选择策略的遗传算法生成测试数据方法. 计算机科学. 2024(06): 95-103 . 百度学术
2.	祁春阳，黄杰，赵翔宇，汪周红. 云边协同的轻量级网络结构人脸识别方法. 东南大学学报(自然科学版). 2023(01): 1-13 . 百度学术
3.	许喆，王志宏，单存宇，孙亚茹，杨莹. 基于重构误差的无监督人脸伪造视频检测. 计算机应用. 2023(05): 1571-1577 . 百度学术
4.	封筠，史屹琛，高宇豪，贺晶晶，余梓彤. 二次解耦与活体特征渐进式对齐的域自适应人脸反欺诈. 计算机研究与发展. 2023(08): 1727-1739 . 本站查看
5.	章育涛，黎英，杨雅莉. 社交网站图像分析研究综述. 信息技术与信息化. 2023(08): 114-121 . 百度学术
6.	史屹琛，封筠，肖立轩，贺晶晶，胡晶晶. 领域外人脸活体检测综述. 计算机科学与探索. 2022(11): 2471-2486 . 百度学术
7.	李书领，魏君飞，庄岩，曹仰杰，李颉，任红军. 基于频域水印的人脸图像窜改检测认证方法. 计算机应用研究. 2022(12): 3776-3780 . 百度学术