Deepfake Detection Using LSTM‐Based Neural Network
Tejaswini Yesugade, Shrikant Kokate, Sarjana Patil, R. Prudhvi Venkata Narasimha Varma, Sejal Pawar · 2022
In this rapid growth of social media, new developments in deep generative network have improve quality of creating more realistic fake videos; such videos are called as deepfake video. Such deepfake videos are used in politics to create political distress, for blackmail and for terrorism events. In order to reduce harm that can be done using such methods and prevent spread of such fake images or videos, we proposed a method that can detect such deepfake. In this paper, we proposed a new method to detect AI-generated fakeIvideos using algorithm such as CNN and LSTM. Our method will detect deepfake by using ResNext50 and LSTM algorithms, which gives accuracy around 88%.