ORIGINAL

Abstract

There has been a wide increase of fake news through different channels, many social media platforms are looking to limit this growth. We propose a new solution to help tackle deepfakes, a popular form of fake news which can present a big challenge to fair reporting due to their ease of spread and their close representation of reality. This solution relies on research on human eyeblink patterns and how they compare to trends seen on deepfaked videos. The results of these studies provide an angle for deepfakes detection that we explore through this project, using computer vision tools.