Paper Title
Deep Fake Detection Using LSTM, Resnetv2, Opencv and Genetic Algorithm
Abstract
Deep learning is the root of the phrase (deep fake). The deep fake is one field that has recently used deep
learning. Due to how simple it is for any user to make fake data using applications and data that are readily available on the
Internet, deep fake poses a danger to privacy and security. Therefore, (detecting deep fakes) is a new technology that has
recently emerged to deal with deep fakes and reduce their severity. In this study, the LSTM and Resnetv2 technology is
combined with genetic algorithms and OpenCV library. In this study, Resnetv2 is used to extract features from movies and
convert them to pictures before passing them on to LSTMs. Then, OpenCV expands and highlights the face based on the
training data. In addition, an animated version of the extracted image will be generated to help detect differences between the
original and edited versions. Consequently, the genetic algorithm will provide a percentage of the training data's best
accuracy.
Keywords - Deep Fake Detection, ResNetV2, LSTM, Genetic Algorithm, OpenCV.