Humans have been generally considered as experts for face identity processing (FIP). However, increasing evidence demonstrates considerable individual differences in behaviorally measured FIP ability - from individuals with profound impairments, to so-called "Super-Recognizers" (SRs). Despite their theoretical and practical implications, e.g. for law enforcement and the judicial system, the sources of FIP ability variation remain elusive. To date \textit{objective and reliable} human FIP ability assessment is critical but lacking, and our understanding of the brain-behavior mechanisms involved is limited. We explored the reliability of frequency-tagged neural correlates of high-level visual perception obtained across considerable time periods using fast periodic visual stimulation (FPVS) EEG in a unique cohort of participants: law enforcement practitioners with documented superior FIP ability. Adopting a deep-data approach, we provide an in-depth report of behavioral and neural data collected from 11 Berlin Police SRs. Specifically, we relate behavioral performance across eight challenging and professionally relevant behavioral FIP tests, to FPVS EEG data for three experiments measuring neural face categorization and face discrimination, measured twice with a 2.5 year re-test interval. Overall, we find medium-to-high test/re-test reliability of SRs' FPVS EEG responses –– both at the group and the individual level, and across experiments. Notably, this contrasts with previous reports of low-to-modest behavioral and neural re-test reliability among neurotypical civilian samples. This seminal report of neural reliability in law enforcement SRs represents a benchmark for future research and practice, in particular high-stakes applied scenarios relying on human (expert, witness) judgments of unfamiliar facial identity.

15 y after their initial discovery, so-called Super-Recognizers (SRs) – individuals with exceptional face recognition abilities – are increasingly deployed by international police agencies. Currently, however, there is lacking empirical evidence supporting their utility for law enforcement. A single study, published 2 y ago, suggests that individuals identified as SRs in the lab can improve forensic perpetrator identification. However, its observations were based on an extremely small forensic stimulus set, obtained from a predominantly civilian sample. Here, reporting data collected among all ~18K Berlin Police officers employed in 2021, we provide the first large-scale, real-world validation of SR deployment, and publicly share all individual-level behavioral data obtained. We reports officers' performance under operationally relevant, highly challenging 1:n facial identity matching conditions via the Berlin Test for Super-Recognizer Identification (beSure®) – the only existing tool that uses authentic forensic material. Our findings show that, as a group, lab-identified SRs exhibit increased proficiency across a range of forensically relevant tasks. However, lab-based SR selection methods showed limited sensitivity at the individual level. This study emphasizes the value of bespoke, professionally relevant assessment for deployment within law enforcement, and the need for empirical investigations into SR deployment. As agencies increasingly integrate biometric technologies, evidence-based approaches for understanding humans' varied perceptual capacities and their implications for decisions is imperative in order to improve criminal investigation and prosecution.

Nearly fifteen years ago, the term Super-Recognizer (SR) was introduced in the scientific literature as a result of an accidental finding. Since then, there has been a surging interest in the topic — both from within the research community, as well as from international media and (mostly European) law enforcement. Given the hype and hope around SRs, a review of our current understanding of their ability is needed. I review findings and outline future avenues for SR work, which extend beyond advancing our understanding of human cognition. The topic “face recognition” attracts attention from various stakeholders, often with opposing views and interests, including e.g. NGOs, policy makers, media, and industry, who are mostly not involved in fundamental research. Therefore, the science around SRs can serve as an entry point to facilitate societally relevant discussions, e.g. around neurodiversity, and our right to privacy and security. This opportunity comes with the responsibility to adopt the highest standards in terms of scientific rigor and transparency, and ensure that high quality, participatory and inclusive research is disseminated honestly and broadly.

The term Super-Recognizer (SR), which describes individuals with supposedly superior facial recognition abilities, may be something of a misnomer. In the same way that blind individuals would not be considered prosopagnosic, SR diagnoses should emphasise at least face identity processing (FIP) specificity, if not recognition in particular. However, SRs tend to be diagnosed with face-specific behavioral tasks, probing either perception and/or recognition, and leaving the neural basis and mechanisms underlying their abilities largely unexplored. The present study therefore sought to determine whether any common FIP subprocesses, among a sample of stringently and comparably diagnosed SRs, would distinguish them from neurotypical controls. To this end, we conducted three Fast Periodic Visual Stimulation (FPVS) EEG experiments in a group of Berlin Police officers identified as SRs using the only existing formal diagnostic framework for lab-based SR identification (Ramon, 2021) that aligns with the seminal study of SRs (Russell et al., 2009). These experiments aimed to isolate FIP from behavioral and general perceptual factors in terms of both the consistency and speed of face identity discrimination and categorization. Broadly, the results of all three experiments provided two key findings. First, whichever factors distinguish SRs from controls, they are not face-specific. Second, SRs are not all cut from the same cloth. Rather, the factors distinguishing SRs from controls seem to be individual-specific, warranting more nuanced and bespoke testing criteria for their deployment in practical applications.

In face processing, an own-age recognition advantage has frequently been reported whereby observers are better at recognizing faces of their own compared to other age groups. We wanted to know whether own-age effects exist in voice recognition. Two listener groups, younger adults (n = 42, 19–35 years, 21 males) and older adults (n = 32, 65–83 years, 14 males), completed a speaker discrimination task (same/different speakers), which included younger and older adult speakers of both sexes. Results revealed no interaction of the factors speaker and listener age and speaker and listener sex on listeners’ sensitivity (d′). Main effects were significant for listener age (young adult listeners exhibited higher sensitivity than the older adult listeners) and speaker sex (listeners’ sensitivity was higher for male compared to female voices). Crucially, response bias (c) revealed that listeners had a significantly higher ‘same’ bias when hearing younger speakers and female speakers. Our findings have implications for theories of voice identity processing and forensic contexts requiring discrimination of speakers’ identity, e.g. earwitnesses telling apart younger and female speakers.

Recent developments in generative AI offer the means to create synthetic identities, or deepfakes, at scale. As deepfake faces and voices become indistinguishable from real ones, they are considered as promising alternatives for research and development to enhance fairness and protect humans’ rights to privacy. Notwithstanding these efforts and intentions, a basic question remains unanswered: Are natural faces and facial deepfakes perceived and remembered in the same way? Using images created via professional photography on the one hand, and a state-of-the-art generative model on the other, we investigated the most studied process of face cognition: perceptual matching and discrimination of facial identity. Our results demonstrate that identity discrimination of natural and synthetic faces is governed by the same underlying perceptual mechanisms: objective stimulus similarity and observers’ ability level. These findings provide empirical support both for the societal risks associated with deepfakes, while also underscoring the utility of synthetic identities for research and development.

Individuals can strongly vary in their ability to process face identity. Understanding the mechanisms driving these differences is important for theoretical development, and in clinical and applied contexts. Here we investigate the role of face-space properties in relation to individual face identity processing skills. We consider two fundamental properties of face-space: expansion (how distant from each other similar faces are located in such space) and adaptability (the degree to which these distances change over time). Fifty-two participants performed a face detection task, with faces systematically varying in their location in face-space, and a comprehensive face identity processing test battery. We replicate previous results indicating a detection advantage for typical, as compared to distinctive faces. Critically, we find that our measure of face-space expansion is not related to individual face processing abilities. However, one measure of face-space adaptability revealed that medium performers exhibit larger adaptation than high or low performers. This pattern seems to be especially driven by a test tapping into long-term memory for faces. While future studies might benefit from the use of more sensitive measures of face-space properties, these results suggest that perceptual adaptation, rather than face-space expansion, may be a contributor to individual differences in face memory.

Super recognizers (SRs) are people that exhibit a naturally occurring superiority for processing facial identity. Despite the increase of SR research, the mechanisms underlying their exceptional abilities remain unclear. Here, we investigated whether the enhanced facial identity processing of SRs could be attributed to the lack of sequential effects, such as serial dependence. In serial dependence, perception of stimulus features is assimilated toward stimuli presented in previous trials. This constant error in visual perception has been proposed as a mechanism that promotes perceptual stability in everyday life. We hypothesized that an absence of this constant source of error in SRs could account for their superior processing—potentially in a domain-general fashion. We tested SRs (n = 17) identified via a recently proposed diagnostic framework (Ramon, 2021) and age-matched controls (n = 20) with two experiments probing serial dependence in the face and shape domains. In each experiment, observers were presented with randomly morphed face identities or shapes and were asked to adjust a face's identity or a shape to match the stimulus they saw. We found serial dependence in controls and SRs alike, with no difference in its magnitude across groups. Interestingly, we found that serial dependence impacted the performance of SRs more than that of controls. Taken together, our results show that enhanced face identity processing skills in SRs cannot be attributed to the lack of serial dependence. Rather, serial dependence, a beneficial nested error in our visual system, may in fact further stabilize the perception of SRs and thus enhance their visual processing proficiency.

The present study is the first empirical investigation of the relationshion between human deepfake detection performance (DDP) and individuals' face identity processing ability. Using videos from the Deepfake Detection Challenge, we investigated DDP in two unique observer groups: Super-Recognizers (SRs) and "normal" officers from within the 18K members of the Berlin Police. SRs were identified either via previously proposed lab-based procedures or the only existing tool for SR identification involving increasingly challenging, authentic forensic material: the Berlin Test For Super-Recognizer Identification (beSure®). Participants judged either pairs of videos, or single videos in a 2-alternative forced-choice decision setting (i.e., which of the pair, or whether a single video was a deepfake or not). We explored speed-accuracy trade-offs, compared DDP between lab-identified SRs and non-SRs, and police officers as a function of their independently measured face identity processing (FIP) ability. Interestingly, we found no relationship between DDP and FIP ability. Further work using static deepfakes created with current state-of-the-art generative models is needed to determine the value of SR deployment for deepfake detection in law enforcement.

Why are some individuals better at recognising faces? Uncovering the neural mechanisms supporting face recognition ability has proven elusive. To tackle this challenge, we used a multi-modal data-driven approach combining neuroimaging, computational modelling, and behavioural tests. We recorded the high-density electroencephalographic brain activity of individuals with extraordinary face recognition abilities—super-recognisers—and typical recognisers in response to diverse visual stimuli. Using multivariate pattern analyses, we decoded face recognition abilities from 1 second of brain activity with up to 80% accuracy. To better understand the mechanisms subtending this decoding, we compared representations in the brains of our participants with those in artificial neural network models of vision and semantics, as well as with those involved in human judgments of shape and meaning similarity. Compared to typical recognisers, we found stronger associations between early brain representations of super-recognisers and mid-level representations of vision models as well as shape similarity judgments. Moreover, we found stronger associations between late brain representations of super-recognisers and representations of the artificial semantic model as well as meaning similarity judgments. Overall, these results indicate that important individual variations in brain processing, including neural computations extending beyond purely visual processes, support differences in face recognition abilities. They provide the first empirical evidence for an association between semantic computations and face recognition abilities. We believe that such multi-modal data-driven approaches will likely play a critical role in further revealing the complex nature of idiosyncratic face recognition in the human brain.

Police use of face recognition technology (FRT) for public safety purposes is a topic that raises several questions. Here, we adopt a multidisciplinary approach to this subject matter. From a scientific standpoint, we outline technical terms and different FRT use-case scenarios, summarize the results of a nationwide survey of the Swiss Police, and introduce a novel human-focused approach, that places Super Recognizers in the center of future FRT. From a legal standpoint, the process of FRT deployment must respect the right for informational self-determination, and its specific features must be clearly defined. We analyze the current Swiss legal basis for FRT deployment across different scenarios, including the fight against hooliganism, surveillance of international borders, and criminal proceedings. Finally, we provide recommendations for desirable legislative revisions of legally permissible FRT usage under consideration of and respect for physical persons’ privacy.

Die polizeiliche Nutzung der Gesichtserkennungstechnologie (FRT) für Zwecke der öffentlichen Sicherheit ist ein Thema, das mehrere Fragen aufwirft. In diesem Beitrag verfolgen wir einen multidisziplinären Ansatz zu diesem Thema. Aus wissenschaftlicher Sicht erläutern wir die technischen Begriffe und die verschiedenen Einsatzmöglichkeiten von FRT, fassen die Ergebnisse einer landesweiten Umfrage bei der Schweizer Polizei zusammen und stellen einen innovativen, menschenzentrierten Ansatz vor, der Super-Recognizer in den Mittelpunkt des zukünftigen FRT-Einsatzes stellt. Aus rechtlicher Sicht muss der Prozess des FRT-Einsatzes das Recht auf informationelle Selbstbestimmung respektieren und seine spezifischen Merkmale müssen klar definiert sein. Wir analysieren die aktuellen schweizerischen Rechtsgrundlagen für den Einsatz von FRT in verschiedenen Szenarien, darunter die Bekämpfung von Hooliganismus, die Überwachung der internationalen Grenzen und die Strafverfolgung. Zuletzt geben wir Empfehlungen für Gesetzesrevisionen, die eine rechtlich zulässige Nutzung von FRT ermöglichen würden, welche die Privatsphäre natürlicher Personen berücksichtigt und respektiert.

L’utilisation de la technologie de reconnaissance faciale (FRT) par la police à des fins de sécurité publique soulève plusieurs questions. Dans cet article, nous adoptons une approche multidisciplinaire. D’un point de vue scientifique, nous définissons les termes techniques et les différentes utilisations possibles de la FRT, nous résumons les résultats d’une enquête nationale menée auprès des corps de police suisses et nous présentons une nouvelle approche centrée sur l’humain, qui place les Super-Recognizers au centre de l’utilisation future de la FRT. D’un point de vue juridique, tout déploiement de la FRT doit respecter le droit à l’autodétermination informationnelle et ses caractéristiques spécifiques doivent être clairement définies. Nous analysons le cadre légal suisse en vigueur pour déterminer les possibilités d’utilisation de la FRT dans différents scénarios, en particulier la lutte contre le hooliganisme, la surveillance des frontières internationales et les procédures pénales. Enfin, nous formulons des recommandations en vue d’une révision législative qui permettrait une utilisation légale de la FRT, qui respecte la vie privée des personnes physiques.

Studies of facial identity processing typically assess perception and/or recognition, with designs differing with respect to one important aspect: Target Prevalence. That is, some include “target absent” (TA) among “target present” (TP) trials. In visual search tasks, TA trials shift an observer’s decisional criterion towards a stricter one, increasing error rates. However, decisional biases will differ inter-individually and can change intra-individually as well. From one standpoint, excluding TA trials is logical as it ensures comparable levels of expectation, or decisional bias across observers, and tasks. However, in reality, TA trials may occur, e.g. in police line-ups, where it is important to consider observers’ face recognition ability independently for TA and TP trials. To our knowledge, the effect of including TA trials has not been systematically investigated in tests of face recognition. We sought to fill this void by testing different versions of the previously established Models Memory Test that measures old/new recognition of experimentally learned facial identities. Our study aimed to answer the open question of whether — and if, how — observer expectation matters in face recognition with naturalistic stimulus variations. We discuss implications for line-up scenarios that are simulated in research settings and occur regularly in policing.

Did you know that some people are really good at remembering and matching faces? They are called super-recognizers. Police want to use super-recognizers to help with their cases. They could sort through camera recordings of crimes and identify criminals. But their abilities have never been tested using real police material.

First we used tests to identify super-recognizers. Then we used real police material to see if super-recognizers could identify criminals. We found that the tests could indeed identify super-recognizers. Super-recognizers did a good job at finding criminals. Much better than regular people! We can use this information to help solve crimes and keep people safe.

We report our ongoing efforts towards the development of challenging speaker discrimination tests. Our goal is to establish highly sensitive tests that enable characterization of individual differences in voice processing skills. Such tools are required but lacking for the identification of so-called “voice super-recognizers” – individuals with superior voice processing ability – who could aid criminal investigations involving audio material. To find such individuals we manipulated test difficulty using an ASR and delta F0-based stimuli selection method. Independent groups of participants performed 2-alternative forced choice voice discrimination tests of stimulus pairs selected either randomly, or systematically based on inter-item similarity. On average, performance was significantly higher for the prior (80.0%) as compared to the latter (68.8%) test. Thus, we have established a method to manipulate task difficulty in speaker discrimination tests. We will further validate and extend our method to include a wider range of audio material implemented as a within-subject design.

About a decade ago, Super-Recognizers (SRs) were first described as individuals with exceptional face identity processing abilities. Since then, various tests have been developed or adapted to assess individuals’ abilities and identify SRs. The extant literature suggests that SRs may be beneficial in police tasks requiring individual identification. However, in reality, the performance of SRs has never been examined using authentic forensic material. This not only limits the external validity of test procedures used to identify SRs, but also claims concerning their deployment in policing. Here, we report the first-ever investigation of SRs’ ability to identify perpetrators using authentic case material. We report the data of 73 SRs and 45 control participants. These include (a) performance on three challenging tests of face identity processing recommended by Ramon (2021) for SR identification; (b) performance for perpetrator identification using four CCTV sequences depicting five perpetrators and police line-ups created for criminal investigation purposes. Our findings demonstrate that the face identity processing tests used here are valid in measuring such abilities and identifying SRs. Moreover, SRs excel at perpetrator identification relative to control participants, with more correct perpetrator identifications, the better their performance across lab tests. These results provide external validity for the recently proposed diagnostic framework and its tests used for SR identification (Ramon,2021). This study provides the first empirical evidence that SRs identified using these measures can be beneficial for forensic perpetrator identification. We discuss theoretical and practical implications for law enforcement, whose procedures can be improved via a human-centric approach centered around individuals with superior abilities. [Accepted manuscript can be accessed here.]

Modern cognitive neuroscience integrates approaches from various fields and, in turn, continues to advance and inspire other domains. Historically, and until relatively recently, single-case and small sample-size approaches dominated neuropsychology and psychophysics, providing in-depth examination of brain-behavior relationships within a limited number of individuals. More recently, technological advances have facilitated our ability to gather and process large amounts of observations, which has, in turn, led to a shift towards "big data” approaches. This has in turn led to a gradual departure from approaches using traditional, modest sample size approaches and a growing trend towards increasing sample sizes in behavioral and neuroimaging studies alike.

The present publication represents the first of an intended collection of reports documenting the registered trademark beSureⓇ and the experience gained in its context. The present report describes the context in which beSureⓇ emerged, as well as its development - from ideation to technical design, up to experimental implementation. On the one hand, this report aims to provide interested practitioners and researchers with information, which is both exhaustive and understandable for readers from all backgrounds. On the other hand, this report also provides methodological documentation required for reproduction and replication of the procedures of beSureⓇ.

Neurotypical observers show large and reliable individual differences in gaze behavior along several semantic object dimensions. Individual gaze behavior towards faces has been linked to face identity processing, including that of neurotypical observers. Here, we investigated potential gaze biases in Super-Recognizers (SRs) - individuals with exceptional face identity processing skills. 10 SRs, identified with a novel conservative diagnostic framework, and 43 controls freely viewed 700 complex scenes, depicting more than 5000 objects. First, we tested whether SRs vs. controls differ in fixation biases along four semantic dimensions: Faces, Text, objects being Touched and Bodies. Second, we tested potential group differences in fixation biases towards eyes and mouths. Finally, we tested whether SRs show less intra- and inter-individual variability with regard to their preferred vertical fixation position in faces. SRs showed a stronger gaze bias towards Faces and away from Text and Touched objects, starting from the first fixation onwards. Further, SRs spent a significantly smaller proportion of first fixations and dwell time towards faces on Mouths but did not differ in dwell time or first fixations devoted to eyes. Face fixation of SRs also fell significantly closer to the theoretical optimal fixation point for identification, just below the eyes. Our findings suggest that reliable superiority for face identity processing is accompanied by early fixation biases towards faces and preferred saccadic landing positions close to the theoretical optimum for face identification. We discuss future directions to investigate the functional basis of individual fixation behavior and face identity processing ability.

This paper aims to provide early-career researchers with a useful introduction to good research practices.

Accurate face identity processing (FIP) is a critical component of security professions. Unfortunately, however, rapid face matching as required in real-life situations such as passport controls cannot be improved via training. While here accuracy is a high priority, it is neither the only, nor most important performance-measure. Officers must process high-throughput information as efficiently as possible – accurately and rapidly. In scenarios with grave public safety implications, however, efficiency is not sufficient. Suspect surveillance and mass-data analysis in criminal investigations also demand processing ample sensitive material consistently over extended periods. Police agencies have sought to optimize operations through personnel selection targeting FIP abilities. Yet to date, the lab-based tests researchers have proffered neither reflect officers’ specific tasks, nor the efficiency and consistency critical to accomplishing them. Therefore, we aimed to benchmark the three most challenging FIP tests available against two work-samples — tasks developed in consultation with police practitioners to measure specific, situationally critical performance. We solicited participation from 390 police officers from Regional Police and Criminal Investigation Departments, yielding a representative sample of 114 participating Protection Police Officers, Mass Data Analysts, and Search Unit Members who regularly employ FIP skills in their work. Data-driven analyses of officers’ FIP abilities revealed that work-sample efficiency and consistency represented most relevant dimensions of variation, and accounted for lab-test performance. Furthermore, performance on either work-sample was better predicted by performance on the other, than by lab-based test scores. This demonstrates the limitations of lab-based tests for applied settings, and stresses the need for predicting police officers’ FIP abilities through contextually and practically relevant performance measures.

In recent years, the number of face identity matching tests in circulation has grown considerably and these are being increasingly utilized to study individual differences in face cognition. Although many of these tests were designed for testing typical observers, recent studies have begun to utilize general-purpose tests for studying specific, atypical populations (e.g., super-recognizers and individuals with prosopagnosia). In this study, we examined the capacity of four tests requiring binary face-matching decisions to study individual differences between healthy observers. Uniquely, we used performance of the patient PS (Rossion, 2018), a well-documented case of acquired prosopagnosia (AP), as a benchmark. Two main findings emerged: (i) PS could exhibit typical rates of accuracy in all tests; (ii) compared to age-matched controls and when considering both accuracy and speed to account for potential trade-offs, only the KFMT — but not the EFCT, PICT or GFMT — was able to detect PS’s severe impairment. These findings reflect the importance of considering both accuracy and response times to measure individual differences in face matching, and the need for comparing tests in terms of their sensitivity, when used as a measure of human cognition and brain functioning.

A face’s memorability refers to the unique combination of its intrinsic visual features facilitating its later recognition. Despite considerable variation in face recognition ability amongst the general population, individuals show substantial concordance regarding the memorability of various faces. And, when the viewpoints across which identities are seen at encoding and recognition differ, such agreement persists, though to a lesser extent. Consequently, face recognition cannot rely solely on image-dependent encoding; individuals must extract some invariant facial information, robust to changes in viewpoint, to do so consistently. However, whether such consistency covaries with overall face processing ability is unclear. Here, therefore, in two experiments we tested recognition of (i) implicitly encoded face images and (ii) explicitly encoded identities in a group of normal control observers against a group of “Super-Recognizers” (SRs) who possess exceptional face processing skills. When implicit encoding was surreptitiously solicited, recognition of studied images was comparable between groups. Yet, when encoding was explicitly solicited, SRs more accurately recognized studied identities across viewpoint changes than normal observers. Critically, image-dependent information could only inform recognition in the first experiment, whereas viewpoint-invariant information could inform recognition consistently in both. Individualized profiles of observers’ performance (as a function of stimulus memorability) reveal that only SRs performed consistently between experiments. We suggest that SRs’ unique capacity for utilizing viewpoint-invariant information for recognition, regardless of encoding conditions, is rooted in fundamentally more accurate and robust representations of identity-based memorability. These results invite a reinterpretation of face memorability that describes viewpoint-invariant information, diagnostic of facial identity representations in memory.

Facial identity matching ability varies widely, ranging from prosopagnosic individuals (who exhibit profound impairments in face cognition/processing) to so‑called super‑recognizers (SRs), possessing exceptional capacities. Yet, despite the often consequential nature of face matching decisions—such as identity verification in security critical settings—ability assessments tendentially rely on simple performance metrics on a handful of heterogeneously related subprocesses, or in some cases only a single measured subprocess. Unfortunately, methodologies of this ilk leave contributions of stimulus information to observed variations in ability largely un(der)specified. Moreover, they are inadequate for addressing the qualitative or quantitative nature of differences between SRs’ abilities and those of the general population. Here, therefore, we sought to investigate individual differences—among SRs identified using a novel conservative diagnostic framework, and neurotypical controls—by systematically varying retinal availability, bandwidth, and orientation of faces’ spatial frequency content in two face matching experiments. Psychophysical evaluations of these parameters’ contributions to ability reveal that SRs more consistently exploit the same spatial frequency information, rather than suggesting qualitatively different profiles between control observers and SRs. These findings stress the importance of optimizing procedures for SR identification, for example by including measures quantifying the consistency of individuals’ behavior.

Best Practice Handbook of the EU-Project SafeCi – Safer Space for Safer Cities.

When you hear the word Super-Recognizer, you may think of comic-book-hero-esque agents searching the underground to find people who went missing decades ago. Compared to this fantasy, the reality seems somewhat less exciting. Super-Recognizers (SRs) were initially reported a decade ago as a collateral while developing tests for developmental prosopagnosia. Today, the topic of SRs sparks interest from groups seeking to enhance scientific knowledge, public safety, or their monetary gain. With no immediate consequences of erroneous SR identification, there has been no pressure to establish a clear SR definition. This promotes heterogenous empirical evidence and the proliferation of unsupported claims in the media. Not only is this status quo unfortunate, it stands in opposition to the potential of special populations – both for science and application. SRs are a special population with imminent real-world value that can advance our understanding of brain functioning. To exploit their potential, I propose a needed formal framework for SR diagnosis, and introduce 70 cases identified based hereupon. These cases represent the core of a growing SR cohort, studied in my lab in the course of a long-term, multi-methodological research agenda involving academic and government collaborators. Finally, I provide recommendations for those interested in SR work, and highlight current caveats and future challenges.

Recognizing individuals based on their face represents a critical component of police work – from identity checks to image intelligence. This capacity, which is at best trainable to a limited extent, varies from person to person and represents one of the greatest challenges for the human brain. In particular processing of unfamiliar faces is extremely error-prone. At the same time, the demand for this task is steadily increasing due to increasing availability of video and image material. To address this, some police agencies are interested in deploying so-called Super-Recognizers: individuals who have a natural propensity for processing faces, and who do so extremely proficiently without any form of training. Super-Recognizers were first scientifically introduced in 2009 – their future lies in the collaboration between police and scientific research.

[Personen anhand ihres Gesichts zu erkennen, stellt einen wesentlichen Bestandteil der polizeilichen Arbeit dar – von der Personenkontrolle bis hin zur Bildfahndung. Diese Fähigkeit, die sich nur bedingt trainieren lässt, variiert von Person zu Person und ist eine der herausforderndsten Aufgaben für das menschliche Gehirn. Vor allem die Verarbeitung unbekannter Gesichter ist extrem fehleranfällig. Gleichzeitig ist dies eine Aufgabe, die aufgrund der Zunahme an zu bearbeitendem Video- und Bildmaterial stetig wächst. Um dieser Situation gerecht zu werden, setzen einige Behörden deshalb auf sogenannte Super-Recognizer: Personen, die von Natur aus und ohne Training Gesichter extrem gut verarbeiten und wiedererkennen können. Super-Recognizer wurden von der Wissenschaft erstmals 2009 thematisiert – die Zukunft liegt in der Zusammenarbeit zwischen Polizei und Forschung.]

Rossion et al. (2020) review over a decade of work investigating the neural basis of unfamiliar face individuation (FI)—the brain's ability to distinguish unfamiliar face identity—using fast periodic visual stimulation (FPVS). Though FPVS measures rapid, automatic processing, its value for studying vision and face cognition could be increased by addressing three important aspects.

Recent investigations of individual differences have demonstrated striking variability in performance both within the same subprocess in face cognition (e.g. face perception), but also between two different subprocesses (i.e. face perception versus face recognition) that are assessed using different tasks (face matching versus face memory). Such differences between and within individuals between and within laboratory tests raise practical challenges. This applies in particular to the development of screening tests for the selection of personnel in real-world settings where faces are routinely processed, such as at passport control. The aim of this study, therefore, was to examine the performance profiles of individuals within and across two different subprocesses of face cognition: face perception and face recognition. To this end, 146 individuals completed four different tests of face matching—one novel tool for assessing proficiency in face perception, as well as three established measures—and two benchmark tests of face memory probing face recognition. In addition to correlational analyses, we further scrutinized individual performance profiles of the highest and lowest performing observers identified per test, as well as across all tests. Overall, a number of correlations emerged between tests. However, there was limited evidence at the individual level to suggest that high proficiency in one test generalized to other tests measuring the same subprocess, as well as those that measured a different subprocess. Beyond emphasizing the need to honour inter-individual differences through careful multivariate assessment in the laboratory, our findings have real-world implications: combinations of tests that most accurately map the task(s) and processes of interest are required for personnel selection.

Unfamiliar face processing is an ability that varies considerably between individuals. Numerous studies have aimed to identify its underlying determinants using controlled experimental procedures. While such tests can isolate variables that influence face processing, they usually involve somewhat unrealistic situations and optimized face images as stimulus material. As a consequence, the extent to which the performance observed under laboratory settings is informative for predicting real-life proficiency remains unclear. We present normative data for two ecologically valid but underused tests of face matching: the Yearbook Test (YBT) and the Facial Identity Card Sorting Test (FICST). The YBT (n = 252) measures identity matching across substantial age-related changes in facial appearance, while the FICST (n = 218) assesses the ability to process unfamiliar facial identity despite superficial image variations. To determine the predictive value of both tests, a subsample of our cohort (n = 181) also completed a commonly used test of face recognition and two tests of face perception (the long form of the Cambridge Face Memory Test (CFMT+), the Expertise in Facial Comparison Test (EFCT) and the Person Identification Challenge Test (PICT)). Focusing on the top performers identified independently per test, we made two important observations: 1) YBT and FICST performance can predict CFMT+ scores and vice versa; and 2) EFCT and PICT scores neither reliably predict superior performance in ecologically meaningful and challenging tests of face matching, nor in the most commonly used test of face recognition. These findings emphasize the necessity for using challenging and ecologically relevant, and thus highly sensitive, tasks of unfamiliar face processing to identify high-performing individuals in the normal population.

This article provides a response to five excellent commentaries on our article ‘Super‐recognizers: From the lab to the world and back again’. Specifically, the response summarizes commonalities between these commentaries. Based on this consensus, we propose a flexible framework for the assessment of superior face recognition and outline guiding principles to advance future work in the field.

The recent discovery of individuals with superior face processing ability has sparked considerable interest amongst cognitive scientists and practitioners alike. These ‘Super‐recognizers’ (SRs) offer clues to the underlying processes responsible for high levels of face processing ability. It has been claimed that they can help make societies safer and fairer by improving accuracy of facial identity processing in real‐world tasks, for example when identifying suspects from Closed Circuit Television or performing security‐critical identity verification tasks. Here, we argue that the current understanding of superior face processing does not justify widespread interest in SR deployment: There are relatively few studies of SRs and no evidence that high accuracy on laboratory‐based tests translates directly to operational deployment. Using simulated data, we show that modest accuracy benefits can be expected from deploying SRs on the basis of ideally calibrated laboratory tests. Attaining more substantial benefits will require greater levels of communication and collaboration between psychologists and practitioners. We propose that translational and reverse‐translational approaches to knowledge development are critical to advance current understanding and to enable optimal deployment of SRs in society. Finally, we outline knowledge gaps that this approach can help address.