RECOGNITION OF PROFILE FRADULENCE FRETTING OUT USING CNN

Abstract

The widespread availability of cameras over the last several decades has made picture capture more common, and the photos we create with these cameras have quickly become an integral part of our everyday lives owing to the wealth of information they hold. Yet, with the proliferation of image-editing software, fabricated photos are increasingly being used to convey disinformation. While there are tried-and-true methods for spotting fakes, recent years have seen a surge of interest in the use of convolutional neural networks (CNNs) for this purpose. But the currently available CNN-based algorithms can only detect certain kinds of forgeries. Therefore, a more effective and precise method of detecting undetected forgeries in a picture is required. In this research, we offer a lightweight deep learning-based system capable of detecting forgeries created using double image compression[1]. Compared to the existing state-ofthe-art methods, our model, which is trained on the difference between the original and compressed versions of a picture, performs far better. Overall validation accuracy of 92.23 percent indicates that the experimental findings are encouraging.