Implementation of a Deepfake Detection System using Convolutional Neural Networks and Adversarial Training

Due to the threat of spreading erroneous information, the employment of deep convolutional neural network (CNN) techniques has resulted in an increase in the number of altered images. A trustworthy and effective mechanism for identifying false photographs is required to solve this problem. In this study, we used the V4D architecture to analyze manipulated videos and concentrated on identifying DeepFake videos in three different contexts: I identifying all types of manipulations, (ii) identifying single manipulations, and (iii) identifying cross manipulation detection, which was used to assess the videos' authenticity. To do this, we used a variety of manipulation techniques and created algorithms to classify undiscovered manipulation methods.