How Close Are Other Computer Vision Tasks to Deepfake Detection?

2023 IEEE INTERNATIONAL JOINT CONFERENCE ON BIOMETRICS, IJCB(2023)

引用 0|浏览7
暂无评分
摘要
In this paper, we challenge the conventional belief that supervised ImageNet-trained backbones have strong generalizability and are suitable for use as feature extractors in deepfake detection models. We present a new measurement, "backbone separability," for visually and quantitatively assessing a backbone's raw capacity to separate data in an unsupervised manner. We also present a systematic benchmark for determining the correlation between deepfake detection and other computer vision tasks using backbones from pre-trained models. Our analysis shows that before fine-tuning, face recognition backbones are more closely related to deepfake detection than other backbones. Additionally, backbones trained using self-supervised methods are more effective in separating deepfakes than those trained using supervised methods. After fine-tuning all backbones on a small deepfake dataset, we found that self-supervised backbones deliver the best results, but there is a risk of overfitting. Our results provide valuable insights that should help researchers and practitioners develop more effective deepfake detection models.
更多
查看译文
关键词
Deepfake Detection,Face Recognition,Unsupervised Manner,Model Selection,Image Classification,Dataset Size,Transfer Learning,Training Methods,ImageNet,Generative Adversarial Networks,Latent Space,Face Images,Model Discrimination,Unseen Data,Self-supervised Learning,Number Of Annotations,Fine-tuned Model,Dimensionality Reduction Algorithms,Face Recognition Task,Poor Separation,Fake Images,Backbone Architecture,High True Positive Rate,Decision Boundary,Age Estimation,Deep Learning,Largest Dataset,Training Dataset,Training Data
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要