Face Morphing Attack Detection Methods

Abstract

Morphing attacks pose a serious threat to face recognition systems, especially in the border control scenario. In order to guarantee a secure operation of face recognition algorithms in the future, it is necessary to be able to reliably detect morphed facial images and thus be able to reject them during enrolment or verification. This chapter provides an overview of morphing attack detection algorithms and metrics to measure and compare their performance. Different concepts of morphing attack detection are introduced and state-of-the-art detection methods are evaluated in a comprehensive cross-database experiments considering various realistic image post-processings.

Improving face morph detection with the pairs training effect

“…However wide-scale implementation across government departments requires significant testing, training, and implementation investments. As discussed previously, there are also significant limitations that need to be overcome in order to develop a robust and reliable algorithm appropriate for implementation across official departments (Scherhag et al, 2022). In the meantime, we propose a simple, quick, and reliable form of training that requires minimal effort to implement.…”

“…As the quality of face morphs increases, they are likely to become even more difficult to detect. A growing body of literature has shown there is potential to develop increasingly sophisticated computer algorithms that can detect face morphs (Makrushin et al, 2017;Neubert, 2017;Raghavendra et al, 2017aRaghavendra et al, , 2017bScherhag et al, 2017Scherhag et al, , 2019Scherhag et al, , 2022Seibold et al, 2018;Venkatesh et al, 2021).…”

“…Often this is a limited set of images that may have been created in the same way. This leads to an overoptimistic detection rate (Scherhag et al, 2022). In addition, the large databases that are publicly accessible are only available to test morph attack detection systems rather than to be used for training the algorithms (Venkatesh et al, 2021).…”

“…The introduction of such algorithms would require a systematic shift in process for the passport office and/or border control. In addition, although a significant volume of research is being dedicated to creating computer algorithms that can reliably detect face morphs, there are some key challenges hindering the development of an algorithm that can produce robust and reliable performance in a real‐life border control or passport issuing office scenario (Ferrara & Franco, 2022; Scherhag et al, 2022; Venkatesh et al, 2021). For example, a lack of availability of large datasets for training, and databases including both digital and print images, means that algorithm developers can only train on what is available to them.…”

A novel blind tamper detection and localization scheme for multiple faces in digital images

“…They described the work in this field as "a cat and mouse game"; whereas the detection techniques are improved, the manipulation methods are improved. Many researches have been introduced tackling the manipulation detection issue; however, to the current date, there is no global reliable face manipulation detection technique, which means this field is still nascent [48][49][50]. One of the limitations that have been mentioned in [48] is the low detection accuracy in the GAN-based methods [51][52][53][54] when bad-quality input images are tested such as images in bad lighting conditions, noisy, blurry, and other low-resolution images.…”