On Detecting GANs and Retouching based Synthetic Alterations
This addresses the issue of synthetic image alterations for face recognition systems, representing a strong specific gain in detection accuracy.
The paper tackles the problem of detecting synthetically altered images, such as those retouched or generated by GANs, which adversely affect face recognition algorithms, and achieves accuracies of 99.65% for retouching detection and 99.83% for distinguishing real images from GAN-generated ones.
Digitally retouching images has become a popular trend, with people posting altered images on social media and even magazines posting flawless facial images of celebrities. Further, with advancements in Generative Adversarial Networks (GANs), now changing attributes and retouching have become very easy. Such synthetic alterations have adverse effect on face recognition algorithms. While researchers have proposed to detect image tampering, detecting GANs generated images has still not been explored. This paper proposes a supervised deep learning algorithm using Convolutional Neural Networks (CNNs) to detect synthetically altered images. The algorithm yields an accuracy of 99.65% on detecting retouching on the ND-IIITD dataset. It outperforms the previous state of the art which reported an accuracy of 87% on the database. For distinguishing between real images and images generated using GANs, the proposed algorithm yields an accuracy of 99.83%.