An Efficient Approach for Image Forgery Detection Using Deep Convolutional Neural Network

The advent of digital platforms revolutionized sharing of personal images and videos on social media. Intruders extract and modify these images to gain popularity or favor. There is freely available software that changes these images very quickly. If the image is forged, the algorithm should detect it and stop to publish on social media. Copy-move and splicing forgery are prevalent methods of image forgery. To combat these forgeries, a real-time deep learning-based approach has been developed, providing enhanced social security in today's society. This approach can effectively detect and identify counterfeit images shared on various social media platforms. The real-time forgery detection algorithms are developed using supervised learning. This algorithm is tested on publicly available datasets such as DVMM, CMFD, and Colombia.