Robust and Generalized DeepFake Detection
This paper focuses on detecting DeepFake videos under three distinct scenarios, which are (i) all manipulation detection, (ii) single manipulation Detection, and then (iii) cross manipulation detection used to test the veracity of the videos.
Abstract
Images that are manipulated are prevalent and are on the spike because of the advancement in deep convolutional neural networks (CNNs) techniques. There have been several concerns regarding the advent spread of false information. There exists a need for a reliable and robust method to detect such fake images. In this paper, analysis was done using the architecture SlowFast in detecting manipulated videos. This paper focuses on detecting DeepFake videos under three distinct scenarios, which are (i) all manipulation detection, (ii) single manipulation detection, and then (iii) cross manipulation detection used to test the veracity of the videos. The manipulation methods and designing algorithms to categorize such unknown manipulation techniques were used.