Detection of Facial Forgery in Digital Images

Abstract

Detection of facial forgery in images and videos is a complex and challenging task. For the last years, different methods have been employed to detect facial forgeries. However, facial forgeries are still very challenging to detect. In this paper, a modern facial forgery detection method is presented which automatically detects facial forgeries in videos and images. The proposed method classifies fake and original videos and images. To classify the images, deep learning model, Inception-ResNet has been adopted. The proposed method is validated on publicly available Deepfake TIMIT dataset and effective results have been reported.