Syn2Real: Forgery Classification via Unsupervised Domain Adaptation

In recent years, image manipulation is becoming increasingly more accessible, yielding more natural-looking images, owing to the modern tools in image processing and computer vision techniques. The task of the identification of forged images has become very challenging. Amongst different types of forgeries, the cases of Copy-Move forgery are increasing manifold, due to the difficulties involved to detect this tampering. To tackle such problems, publicly available datasets are insufficient. Addressing this issue, we employed unsupervised domain adaptation to learn the discriminative features from a large dataset and classify the forged images in new domains by feature space mapping. We synthesized a forgery dataset using image inpainting and copy-move forgery algorithm. However, models trained on these synthetic datasets have a significant drop in performance when tested on more realistic data. We improvised the F1 score on CASIA and CoMoFoD dataset to 80.3% and 78.8%, respectively outperforming state-of-the-art copy-move classification algorithms. Our approach can be helpful in those cases where the classification of data is unavailable.