Find the Differences: Differential Morphing Attack Detection vs Face Recognition
For face recognition system deployers, this work provides a practical method to repurpose existing FR systems for morphing attack detection without additional hardware or training.
The paper argues that face recognition (FR) and differential morphing attack detection (D-MAD) perform similar tasks, showing that current FR thresholds cause vulnerability to morphing attacks. It proposes using existing FR systems for morphing detection with a new threshold that limits vulnerability to unknown attacks.
Morphing is a challenge to face recognition (FR) for which several morphing attack detection solutions have been proposed. We argue that face recognition and differential morphing attack detection (D-MAD) in principle perform very similar tasks, which we support by comparing an FR system with two existing D-MAD approaches. We also show that currently used decision thresholds inherently lead to FR systems being vulnerable to morphing attacks and that this explains the tradeoff between performance on normal images and vulnerability to morphing attacks. We propose using FR systems that are already in place for morphing detection and introduce a new evaluation threshold that guarantees an upper limit to the vulnerability to morphing attacks - even of unknown types.