A Study on Analyzing Deepfake through various Facial Regions- A Review

In today's scenario, Deepfakes images and videos are being spread globally in social media websites like facebook, twitter and instagram to divert the mindset of people and frame the particular opinion. Many android applications like Faceapp are using Generative Adversarial Network (GAN) to imitate realistic images. Moreover, some free deepfake android app such as Zao, The Chinese app that help users to interchange faces in an easy manner with well-known personalities of the film industry so they can look as if they are being casted for the role in the movie, There are great requirements for detecting deepfakes due to the holistic misuse. Although many deepfakes detection algorithms are already existing in this aspect, very little work has been done to detect deepfake on the basis of analyzing normal eye blink rate. This paper presents a review on detecting deepfake through analyzing various facial regions like normal eye blink rate of a human being because it is not generally feasible to detect eye blink rate of a person at normal rate due to the deficiency of closed eye images.