ADVANCING VIDEO FORGERY DETECTION VIA DEEP CONVOLUTIONAL NEURAL NETWORKS

Abstract

The increasing sophistication of editing technologies has raised concerns about video forgeries and their potential impact on digital forensics, media integrity, and security. It is a challenging task as tiny changes are so difficult to identify with traditional detection techniques. This paper's main objective is to investigate how Deep Convolutional Neural Networks (DCNN) might enhance the detection of video counterfeiting. The suggested model can identify instances of deepfake alterations, splicing, and frame tampering with an impressive level of accuracy according to deep learning techniques. The trials' findings suggest that deep convolutional neural networks (DCNN) perform better than more conventional methods, which may make them useful in forensic investigations.