Training Strategies and Data Augmentations in CNN-based DeepFake Video Detection

TL;DR

This paper investigates how various training strategies and data augmentation methods influence the performance and robustness of CNN-based deepfake video detectors across different datasets.

Contribution

It provides a comprehensive analysis of training and augmentation techniques to improve deepfake detection accuracy and generalization.

Findings

Certain data augmentations improve cross-dataset performance.

Training strategies significantly affect detector robustness.

Some methods enhance detection accuracy on the training dataset.

Abstract

The fast and continuous growth in number and quality of deepfake videos calls for the development of reliable detection systems capable of automatically warning users on social media and on the Internet about the potential untruthfulness of such contents. While algorithms, software, and smartphone apps are getting better every day in generating manipulated videos and swapping faces, the accuracy of automated systems for face forgery detection in videos is still quite limited and generally biased toward the dataset used to design and train a specific detection system. In this paper we analyze how different training strategies and data augmentation techniques affect CNN-based deepfake detectors when training and testing on the same dataset or across different datasets.