Authors: A.BABISHA
Abstract: With the rapid penetration of the Internet into every part of our daily life, it is agreed that it will be an important medium for future communication, perhaps even more important than the television This product is a self-contained product made to facilitate the users with the facility to detect which video amongst the 2 is a real or fake one. This can be very helpful the society to control and reduce blackmailing and sharing of obscene content. We extract the feature points from the images in the training dataset using FAST and get the feature point descriptors using BRIEF. Then using DLIB face detector to detect face region and regions inside the face. We group the feature points based on the region that they are falling in. Then the feature point descriptors are aggregated to train the random forest classifier.
DOI: http://doi.org/
