Key points are not available for this paper at this time.
Deepfake videos seriously put at risk their trustworthiness and reputation of individuals and public figures. Through the development of Generative Adversarial Networks (GANs), is an advance method to identify the deepfake technologies, it has emerged, causing serious problems for the veracity of graphics. The present investigation discusses the methods for detecting deepfake face swapping in photos and videos using Generative Adversarial Networks. We investigate the use of GANs for face swap using a curated CelebA dataset. We also address in great detail about the difficulties, prevailing theories, and future directions in deepfake technology research. GAN architecture consists of a generator and a discriminator to investigate the fake face. The accuracy of GAN will improve if the number of Epoch increases, on the basis of this we have calculated the loss of Discriminator 0.70558 and Generator loss 0.72143, Epoch 3, iteration 3028, duration 3.928.
Mall et al. (Fri,) studied this question.