Manipulating faces for identity theft via morphing and deepfake: Digital privacy

Digital face images can be easily manipulated for obfuscating or impersonating an identity. Several techniques are used for face manipulation, both traditional computer vision based such as morphing, and modern deep learning based such as deepfake. Morphing and deepfake techniques became advanced enough in creating photorealistic face images. Due to that, these techniques pose a serious threat to identity theft and can significantly harm at a personal level such as the risk of reputation and money, and the national level such as interference in the election. In this chapter, we review (i) different stealthy ways of identity threat generation techniques, (ii) popular databases used in this research direction, and (iii) defense algorithms build to detect these manipulated images. We further provide key open challenges which need to be addressed to make the defense algorithms robust, generalized, and to handle the adaptive nature of the attacks.