A Comparative Analysis of Single Image-Based Face Morphing Attack Detection

Face morph attack detection (MAD) is fusing two faces together a new face image that combines characteristics from both individuals. Both individuals are verified by a biometric facial recognition technology using this image from a passport. The two categories of face morphing attack detection methods consist of (i) single image-based MAD (S-MAD) and (ii) differential image-based MAD (D-MAD). This recent survey mainly focuses on single image-based MAD (SMAD). The single image-based MAD problem is more difficult than the differential image-based MAD task. This paper discusses several approaches, techniques, and databases that are provided in the comparative manner.