DEEPFAKE OR REAL IMAGE PREDICTION USING MESONET

Abstract

Advance development in Machine Learning, Deep Learning, and Artificial Intelli gence (AI) allow people to exchange the faces and voices of other people in videos so that they look like people did or wanted to say. These videos and photos are called ”deepfake” and each day they are more complicated, which worries legislators. This technology uses machine learning technology to provide the computer with real data about the image so that we can falsify it. The creators of Deepfake use artificial in telligence and machine learning algorithms to mimic the work and characteristics of the real human. It differs from traditional fake media because it is difficult to iden tify. As the 2020 election approaches, 4,444 AI-generated DeepFakes have entered the news cycle. DeepFakes threatens facial recognition and online content. This hoax can be dangerous, because if used incorrectly, you can abuse this technique. Fake video, voice and audio clips can cause enormous damage.We will use Mesonet To make predictions on image data. we will examine four sets of images-correctly identified deepfakes, correctly identified reals, misIdentified deepfakes, misIdentified reals and we will see whether the human eye can pick up on any insights into the world of deepfakes.We will be using the Meso 4 model Trained on the deepfake and real data set.