Detection of Deepfake Videos Using Transfer Learning

Deep learning algorithms have become so potent due to increased computing power that it is now relatively easy to produce human-like synthesised videos, sometimes known as “deep fakes.” One may easily imagine scenarios where these realistic face swap deep fakes are used for negative purposes. In this paper, we provide a novel deep learning-based strategy for the efficient separation of fraudulent videos produced by AI from real videos. Automatically spotting replacement and recreation deep fakes is possible with our technology. To combat artificial intelligence, we are attempting to deploy artificial intelligence (AI). The frame-level features are extracted by our system using a Res-Next Convolution neural network, and these features are then used to train an LSTM-based recurrent neural network (RNN) to determine whether the video has been altered in any way or not, i.e. whether it is a deep fake or authentic video. We test our technique on a sizable, balanced, and mixed data set Deepfake detection challenge[1], in order to simulate real-time events and improve the model’s performance on real-time data. We also demonstrate a very straightforward and reliable approach that allows our system to produce results that are competitive.