Abstract
The Oxford University Press book entitled Forensic face matching: Research and practice edited by Markus Bindemann is a comprehensive overview of pertinent issues surrounding the field of face matching. The main strength of this contribution is the combination of theoretical accounts looking at various aspects of face matching, such as factors affecting performance, individual differences, and steps towards a Cognitive Theory of Unfamiliar Face Matching with legal, national security, and forensic applications of this ubiquitous task. The book also covers important contemporary aspects, such as disguises (e.g. hyper-realistic face masks) and automated face recognition (AFR).
The book spans ten chapters which can be broadly split into three themes: Theoretical Aspects of face matching in humans and AFR (chapters 2, 3, 6, 9), Practitioner Perspectives (chapters 1, 7, 8) and Bridging the Gap between the lab and the real- world (chapters 4, 5, 10).
In the Theoretical Aspects of face matching in humans and AFR, Fysh (chapter 2) provides an excellent review of factors limiting performance in face matching from person characteristics (e.g. expression, appearance changes over time, or disguise attempts) through image characteristics (e.g., image quality and lens distortion) to other situational factors (e.g., trial number and live-to-photo comparison). One aspect generally missing from this area of research is the use of make-up (e.g., Mileva, Bobak, & Hancock, submitted), or popular cosmetic procedures, such as botulin injections or derma-fillers. The use of derma-fillers, in particular, can considerably change someone's appearance and it is unclear how they influence face matching.
Further, Bindemann and Burton, and Bate, Mestry, and Porch (chapters 3 and 6, respectively) make important steps towards providing a theoretical foundation for face matching, both in terms of cognitive theory and individual differences in performance. I find the addition of decisional factors to the proposed cognitive theoretical model particularly enthusing. It is likely that decision making at the individual level is as important as the face itself, and the observer's ability (for a comprehensive overview of individual differences in face matching ability see Chapter 6 of this volume), in the final outcome. For example, research from our lab showed that changes in hue of the images in face matching pairs can influence participants’ bias, but not accuracy (Bobak et al., 2019). The decisional bias are widely researched in the eye-witness literature (e.g., Baldassari et al., 2019) and they could have a similarly important place in face matching research.
Although mentioned in passing before, it is only in Chapter 9 that Noyes and Hill really introduce the reader to the relatively new addition in face matching research, namely automated face recognition (AFR). This chapter is particularly helpful by providing a clear and understandable overview of the workings of algorithms, with example tasks where AFR might be employed and rationale behind choosing a specific criterion for an AFR system. Future editions of this volume may wish to discuss the ethics and the citizens’ attitudes to AFR in more detail as with increased applications of AFR, important research regarding attitudes is emerging (e.g., Ritchie et al., 2021).
In the Practitioner Perspectives, Stevens (chapter 1) reviews person identification at airports with some interesting examples of historical and contemporary document design, and the use of e-gates giving an example of human-computer interaction. The chapter also offers some insight into selection, training, and monitoring of passport-control staff. Interestingly, the selection process does not mention basic assessment of the control staff's face matching ability which varies greatly in the general population (as reviewed in detail in chapter 6), including national security staff (White et al., 2014).
In chapter 7, Moreton takes the reader through various techniques in forensic face matching, while in chapter 8, Roberts provides a legal perspective on forensic face matching with respect to admissibility of evidence. Moreton's review of forensic examiners’ work sheds some light on the differences between their approach and face matching performed by so-called super-recognisers, while Roberts details the criteria for establishing reliability of an expert in court.
Chapters 4 and 5 by White, Towler and Kemp, and Towler, Kemp, and White, respectively, bridge the gap between the lab and the real-world perspectives by using theoretical knowledge to understand practitioners’ performance and reviewing training practices aimed at improving forensic face matching. This close collaboration between the lab and the real-world is paramount to progressing the laboratory research in the way that can have positive impact on Society (see also Ramon et al., 2019).
In the concluding chapter, Sanders and Jenkins (chapter 10) describe the threat, increasing incidence, and the work surrounding the use of hyper-realistic face masks as a means to disguise identity. Importantly, this section not only reviews the research, but provides some very practical guidelines for improving detection of hyper-realistic masks that are likely to be helpful to national security employes undertaking identity verification.
In writing this Oxford University Press book, the authors combine their detailed theoretical knowledge and “real-world” experience to provide a much-needed laboratory and practitioner perspective on forensic face matching. Through easy-to-navigate sections with detailed, and yet non-presumptuous (of any background knowledge) content, Bindemann has edited a highly accessible and comprehensive resource for academics, practitioners, and graduate students with interest in all aspects of forensic face matching and visual perception in general.
