Paper Key : IRJ************581
Author: Ritika Sonawane,Sonali Gade,Priyanka Harnawal,Anushka Mandlik
Date Published: 03 Jan 2025
Abstract
This research presents a deep learning model for detecting deepfake images generated by Generative Adversarial Networks (GANs). Our approach utilizes GAN-generated features to train a highly accurate classification model, incorporating advanced techniques like convolutional neural networks and attention mechanisms to enhance performance. The model demonstrates high precision in identifying manipulated images, contributing to the development of countermeasures against the misuse of deepfake technology.The criticality of the issue lies in its potential to weaken believe in advanced media, particularly in touchy areas such as legislative issues, news, and amusement. As the quality of deepfakes proceeds to make strides, it gets to be progressively troublesome for current detection strategies to keep pace. In this way, theres a squeezing require for more viable arrangements that can identify deepfakes in real-time. This project focuses exclusively on leveraging Generative Antagonistic Systems (GANs) to make a capable, adaptable location framework that can relieve the dangers postured by deepfake media.