Enhancing Digital image forgery detection using transfer learning

In recent times, digital images have become widely used across various sectors, making them a prime target for manipulation by malicious actors.The act of altering digital images for fraudulent purposes is referred to as image forgery.To address this challenge, this project proposes an efficient framework for detecting image forgeries using transfer learning models.The findings from this study demonstrate that the proposed method achieves exceptional accuracy in detecting forgeries within a dataset of 220 images, comprising both genuine and manipulated samples.The models utilized in the framework, such as ResNet50, VGG16, Xception, DenseNet201, Mobile Net, and Dense Net Mobile, showcase their effectiveness in accurately identifying tampered images.