FAKE IMAGE IDENTIFICATION USING MACHINE LEARNING

FAKE IMAGE IDENTIFICATION USING MACHINE LEARNING

¹ Mohammed Azmatullah ² Sykam Hemanth ³ Ummadi Arun Vineel raj ⁴ G Durvasi Kiran

^1,2,3 Department of Information Technology, Andhra Loyola Institute of Engineering and Technology

⁴ Associate professer, Department of Information Technology, Andhra Loyola Institute of Engineering and Technology

Abstract:

Biometric techniques are now helpful in identifying a person's identity, but criminals alter their look, behaviour, and psychological makeup to trick identification systems. We are employing a novel method called Deep Texture Features extraction from photos to solve this issue, followed by the construction of a machine learning model using the CNN (Convolution Neural Networks) algorithm. This method is also known as LBPNet or NLBPNet since it largely relies on the LBP (Local Binary Pattern) algorithm for features extraction.

LBPNET, a machine learning convolution neural network, is the name of the network we created for this research to identify fraudulent face photographs. Here, we will first extract LBP from the photos before training the convolution neural network on the LBP descriptor images to create a training model. Every time we submit a new test picture, the training model will be applied to that test image to determine if it includes a false image or not. Details regarding LBP are shown below.

keywords: Biometry, Identity, Recognition, Detection, Fake face.