CONVOLUTIONAL NEURAL NETWORK FOR FAKE FACE DETECTION: A DEEP LEARNING MODEL (KEY IJP************147)

• Chetan Raut,Saloni Chintanwar,Prof.p.s.sontakke

Abstract

The emergence of deepfake technology, which utilizes advanced deep learning techniques to produce highly realistic fake videos, poses significant challenges to the integrity of digital media. This study aims to develop a robust system for detecting deepfake faces in video content by combining Convolutional Neural Networks (CNNs) with Vision Transformers. The system focuses on analyzing both spatial and temporal inconsistencies across video frames to effectively identify manipulated media. To enhance its precision and practical utility, the detection model will be trained using extensive datasets, including FaceForensics++, the DeepFake Detection Challenge (DFDC), and Celeb-DF.