Find the Differences: Differential Morphing Attack Detection vs Face Recognition

TL;DR

This paper compares face recognition and differential morphing attack detection, highlighting their similarities, vulnerabilities due to decision thresholds, and proposing a new threshold to limit morphing attack vulnerability.

Contribution

It demonstrates the similarity between face recognition and D-MAD, analyzes threshold impacts, and introduces a new evaluation threshold to reduce morphing attack vulnerability.

Findings

FR systems are vulnerable to morphing attacks due to decision thresholds.

Current thresholds cause a tradeoff between normal image performance and attack vulnerability.

A new threshold can limit vulnerability even for unknown morphing attack types.

Abstract

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.