Face Morphing Detection using Generative Adversarial Networks
Xinrui Yu, Guojun Yang, Jafar Saniie · 2019
Facial recognition system is used by various entities, including social media, identity verification, and security services. However, face morphing techniques present an adversarial challenge to the current facial recognition system. This is particularly the case for face morphing using Generative Adversarial Networks (GAN) method. To combat the face morphing we studied the GAN to detect morphed facial images. This has been achieved by using GAN to generate a large number of morphed images. Then, the morphed images are used to retrain the GAN for detecting the morphed images. The performance of this method is evaluated using facial images dataset and network structures.