Deep image composition meets image forgery

Abstract Image forgery is a topic that has been studied for many years. Advances in deep learning have impacted image forgery detection as much as they have impacted other areas of computer vision and have improved the state of the art. Deep learning models require large amounts of labeled data for training. In the case of image forgery, labeled data at the pixel level is a very important factor for the models to learn. None of the existing datasets have sufficient size, realism and pixel-level labeling at the same time. This is due to the high cost of producing and labeling quality images. It can take hours for an image editing expert to manipulate just one image. To bridge this gap, data generation is automated using image composition techniques that are very related to image forgery. Unlike other automated data generation frameworks, state of the art image composition deep learning models are used to generate spliced images close to the quality of real-life manipulations. Finally, the generated dataset was tested on the SOTA image manipulation detection model and show that its prediction performance is lower compared to existing datasets, i.e. realistic images that are more difficult to detect were produced. Dataset is available at https://github.com/99eren99/DIS25k .