David White
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
David White has a total of 148 co-authors. Here are the top co-authors by number of shared publications:
- Richard I. Kemp
School of Psychology, University of New South Wales
26 shared publications - James Daniel Dunn
School of Psychology UNSW Sydney Sydney New South Wales Australia
19 shared publications - Alice Towler
UNSW Sydney, Sch Psychol, Mathews Bldg, Kensington, NSW 2052, Australia
18 shared publications - A. Mike Burton
Department of Psychology, University of York, UK
16 shared publications Rob Jenkins
Department of Psychology, University of York, York, UK
10 shared publicationsAlice Towler
School of Psychology, University of Queensland, Queensland 4072, Australia
5 shared publications
Latest Publications
- Open Access
Naturalistic face-directed attention predicts face recognition ability
Recognising people precedes and informs most social interactions, yet it is not clear why individuals differ so markedly in face recognition ability. Exposure-based accounts propose that it is shaped by visual experience with faces—known as ‘face diet’—but most prior work treats exposure as passively acquired, thereby masking the active and dynamic nature of visual attention in natural settings. Here we examined whether spontaneous attention to faces in everyday environments predicts face recognition ability. Eighty participants wore mobile eye-tracking glasses while walking through a busy university campus and during a face-to-face interaction, before completing standard measures of face recognition ability. Fixations directed to faces showed stable individual differences in both settings. People who looked more at faces scored higher on our composite measure of face recognition ability, and gaze in the two settings predicted ability independently, together accounting for 22% of its variance. These findings suggest that differences in how people actively sample visual information from their environments may be an overlooked source of variance in perceptual expertise.
Open AccessNaturalistic face-directed attention predicts face recognition ability
Recognising people precedes and informs most social interactions, yet it is not clear why individuals differ so markedly in face recognition ability. Exposure-based accounts propose that it is shaped by visual experience with faces—known as ‘face diet’—but most prior work treats exposure as passively acquired, thereby masking the active and dynamic nature of visual attention in natural settings. Here we examined whether spontaneous attention to faces in everyday environments predicts face recognition ability. Eighty participants wore mobile eye-tracking glasses while walking through a busy university campus and during a face-to-face interaction, before completing standard measures of face recognition ability. Fixations directed to faces showed stable individual differences in both settings. People who looked more at faces scored higher on our composite measure of face recognition ability, and gaze in the two settings predicted ability independently, together accounting for 22% of its variance. These findings suggest that differences in how people actively sample visual information from their environments may be an overlooked source of variance in perceptual expertise.
- Open Access
- 19 February 2026
- Published by Center for Open Science in PsyArXiv
Too good to be true: Synthetic AI faces are more average than real faces and super-recognizers know it
The AI revolution has produced synthetic faces that often appear more human than photos of real people. We tested whether individual differences in human face recognition ability explain variation in discriminating AI from real faces. Super-recognizers – people with exceptional ability to recognize human faces (N = 36) – outperformed a typical sample by 15% and by 7% compared to a group of higher performing, motivated control participants (Cohen's d = 0.55; N = 89). Individual difference analysis revealed that this pattern reflected a positive association between human face recognition and AI face discrimination abilities. AI discrimination ability was also associated with individuals' sensitivity to the ‘hyper-average’ appearance of AI faces. Deep neural networks optimized for face identity processing confirmed a more central distribution of AI faces in face-space. Moreover, centrality was associated with a higher probability of super-recognizers judging the faces as AI, but this pattern was not observed for controls. Super-recognizers' correct interpretation of hyper-averageness as a cue to artificiality constitutes the first mechanistic link between evolved expertise in face processing and AI face detection and addresses a common misconception regarding the structure of human face space.
- 19 February 2026
- Published by Center for Open Science in PsyArXiv
Open AccessToo good to be true: Synthetic AI faces are more average than real faces and super-recognizers know it
The AI revolution has produced synthetic faces that often appear more human than photos of real people. We tested whether individual differences in human face recognition ability explain variation in discriminating AI from real faces. Super-recognizers – people with exceptional ability to recognize human faces (N = 36) – outperformed a typical sample by 15% and by 7% compared to a group of higher performing, motivated control participants (Cohen's d = 0.55; N = 89). Individual difference analysis revealed that this pattern reflected a positive association between human face recognition and AI face discrimination abilities. AI discrimination ability was also associated with individuals' sensitivity to the ‘hyper-average’ appearance of AI faces. Deep neural networks optimized for face identity processing confirmed a more central distribution of AI faces in face-space. Moreover, centrality was associated with a higher probability of super-recognizers judging the faces as AI, but this pattern was not observed for controls. Super-recognizers' correct interpretation of hyper-averageness as a cue to artificiality constitutes the first mechanistic link between evolved expertise in face processing and AI face detection and addresses a common misconception regarding the structure of human face space.
- Top Cited PapersOpen Access
- 18 February 2026
- Published by Wiley in British Journal of Psychology
- Vol. 117 (4) , 1279-1294
Too good to be true: Synthetic AI faces are more average than real faces and super‐recognizers know it
The AI revolution has produced synthetic faces that often appear more human than photos of real people. We tested whether individual differences in human face recognition ability explain variation in discriminating AI from real faces. Super‐recognizers – people with exceptional ability to recognize human faces ( N = 36) – outperformed a typical sample by 15% and by 7% compared to a group of higher performing, motivated control participants (Cohen's d = 0.55; N = 89). Individual difference analysis revealed that this pattern reflected a positive association between human face recognition and AI face discrimination abilities. AI discrimination ability was also associated with individuals' sensitivity to the ‘hyper‐average’ appearance of AI faces. Deep neural networks optimized for face identity processing confirmed a more central distribution of AI faces in face‐space. Moreover, centrality was associated with a higher probability of super‐recognizers judging the faces as AI, but this pattern was not observed for controls. Super‐recognizers' correct interpretation of hyper‐averageness as a cue to artificiality constitutes the first mechanistic link between evolved expertise in face processing and AI face detection and addresses a common misconception regarding the structure of human face space.
- 18 February 2026
- Published by Wiley in British Journal of Psychology
- Vol. 117 (4) , 1279-1294
Top Cited PapersOpen AccessToo good to be true: Synthetic AI faces are more average than real faces and super‐recognizers know it
The AI revolution has produced synthetic faces that often appear more human than photos of real people. We tested whether individual differences in human face recognition ability explain variation in discriminating AI from real faces. Super‐recognizers – people with exceptional ability to recognize human faces ( N = 36) – outperformed a typical sample by 15% and by 7% compared to a group of higher performing, motivated control participants (Cohen's d = 0.55; N = 89). Individual difference analysis revealed that this pattern reflected a positive association between human face recognition and AI face discrimination abilities. AI discrimination ability was also associated with individuals' sensitivity to the ‘hyper‐average’ appearance of AI faces. Deep neural networks optimized for face identity processing confirmed a more central distribution of AI faces in face‐space. Moreover, centrality was associated with a higher probability of super‐recognizers judging the faces as AI, but this pattern was not observed for controls. Super‐recognizers' correct interpretation of hyper‐averageness as a cue to artificiality constitutes the first mechanistic link between evolved expertise in face processing and AI face detection and addresses a common misconception regarding the structure of human face space.
- Open Access
- 5 November 2025
- Published by Center for Open Science in OSF Preprints
Super-recognisers sample visual information of superior computational value for facial recognition
Super-recognisers—individuals with exceptionally high face recognition abilities—are a key exemplar of biological visual expertise. Recent eye-tracking evidence suggests that their expertise may be driven by exploratory viewing behaviour during learning, but it remains unclear whether this perceptual sampling is functional for face identity processing. Here, we develop a novel approach to quantify the computational value of face information samples and test the utility of information sampling in super-recognisers. Using measurements of eye gaze behaviour, we reconstructed the retinal information that participants acquired while learning new faces. We then evaluated the computational value of this information for face identity processing using nine deep neural networks (DNNs) optimised for this task. Identity matching accuracy improved across all DNNs when using visual information sampled by super-recognisers compared to typical viewers. Interestingly, this advantage could not be explained by the greater quantity of information alone, and so differences in both the quantity and quality of face information encoded on the retina contribute to individual differences in face processing ability. These findings support accounts of visual expertise that emphasise attentional mechanisms and the role of active visual exploration in learning.
- 5 November 2025
- Published by Center for Open Science in OSF Preprints
Open AccessSuper-recognisers sample visual information of superior computational value for facial recognition
Super-recognisers—individuals with exceptionally high face recognition abilities—are a key exemplar of biological visual expertise. Recent eye-tracking evidence suggests that their expertise may be driven by exploratory viewing behaviour during learning, but it remains unclear whether this perceptual sampling is functional for face identity processing. Here, we develop a novel approach to quantify the computational value of face information samples and test the utility of information sampling in super-recognisers. Using measurements of eye gaze behaviour, we reconstructed the retinal information that participants acquired while learning new faces. We then evaluated the computational value of this information for face identity processing using nine deep neural networks (DNNs) optimised for this task. Identity matching accuracy improved across all DNNs when using visual information sampled by super-recognisers compared to typical viewers. Interestingly, this advantage could not be explained by the greater quantity of information alone, and so differences in both the quantity and quality of face information encoded on the retina contribute to individual differences in face processing ability. These findings support accounts of visual expertise that emphasise attentional mechanisms and the role of active visual exploration in learning.
- Open Access
- 5 November 2025
- Published by The Royal Society in Proceedings Of The Royal Society B-Biological Sciences
- Vol. 292 (2058) , 20252005
Super-recognizers sample visual information of superior computational value for facial recognition
Super-recognizers—individuals with exceptionally high face recognition abilities—are a key exemplar of biological visual expertise. Recent eye-tracking evidence suggests that their expertise may be driven by exploratory viewing behaviour during learning, but it remains unclear whether this perceptual sampling is functional for face identity processing. Here, we develop a novel approach to quantify the computational value of face information samples and test the utility of information sampling in super-recognizers. Using measurements of eye gaze behaviour, we reconstructed the retinal information that participants acquired while learning new faces. We then evaluated the computational value of this information for face identity processing using nine deep neural networks (DNNs) optimized for this task. Identity matching accuracy improved across all DNNs when using visual information sampled by super-recognizers compared with typical viewers. Interestingly, this advantage could not be explained by the greater quantity of information alone, and so differences in both the quantity and quality of face information encoded on the retina contribute to individual differences in face processing ability. These findings support accounts of visual expertise that emphasize attentional mechanisms and the role of active visual exploration in learning.
- 5 November 2025
- Published by The Royal Society in Proceedings Of The Royal Society B-Biological Sciences
- Vol. 292 (2058) , 20252005
Open AccessSuper-recognizers sample visual information of superior computational value for facial recognition
Super-recognizers—individuals with exceptionally high face recognition abilities—are a key exemplar of biological visual expertise. Recent eye-tracking evidence suggests that their expertise may be driven by exploratory viewing behaviour during learning, but it remains unclear whether this perceptual sampling is functional for face identity processing. Here, we develop a novel approach to quantify the computational value of face information samples and test the utility of information sampling in super-recognizers. Using measurements of eye gaze behaviour, we reconstructed the retinal information that participants acquired while learning new faces. We then evaluated the computational value of this information for face identity processing using nine deep neural networks (DNNs) optimized for this task. Identity matching accuracy improved across all DNNs when using visual information sampled by super-recognizers compared with typical viewers. Interestingly, this advantage could not be explained by the greater quantity of information alone, and so differences in both the quantity and quality of face information encoded on the retina contribute to individual differences in face processing ability. These findings support accounts of visual expertise that emphasize attentional mechanisms and the role of active visual exploration in learning.
- Open Access
- 29 August 2025
- Published by Center for Open Science in PsyArXiv
Using face averages to measure differential accuracy for demographic groups in facial recognition
Facial recognition Deep Neural Networks (DNNs) and humans both show systematic differences in face recognition accuracy for different demographic groups. Differential accuracy has been shown due to race, age and gender, raising important questions about the impact of facial recognition decisions on the fairness of society. Current methods for measuring bias require curating large databases of face images, which is labor-intensive, producing bespoke tests that cannot be easily shared or standardized due to data privacy. Here we develop a novel solution to this problem inspired by psychological research in face perception. We ask whether face averages can be used to predict face recognition accuracy differentials for DNNs and humans. We generate sets of average images from random samples of exemplar faces from a demographic group to measure the density of these sets in representational space. We find that this approach provides reliable predictions of differential accuracy across demographic groups in both DNNs and human participants. However, we also find evidence that face averages were not represented in the center of face categories. This finding should be addressed in future development of our approach, and also challenges influential cognitive models of face identity representation.
- 29 August 2025
- Published by Center for Open Science in PsyArXiv
Open AccessUsing face averages to measure differential accuracy for demographic groups in facial recognition
Facial recognition Deep Neural Networks (DNNs) and humans both show systematic differences in face recognition accuracy for different demographic groups. Differential accuracy has been shown due to race, age and gender, raising important questions about the impact of facial recognition decisions on the fairness of society. Current methods for measuring bias require curating large databases of face images, which is labor-intensive, producing bespoke tests that cannot be easily shared or standardized due to data privacy. Here we develop a novel solution to this problem inspired by psychological research in face perception. We ask whether face averages can be used to predict face recognition accuracy differentials for DNNs and humans. We generate sets of average images from random samples of exemplar faces from a demographic group to measure the density of these sets in representational space. We find that this approach provides reliable predictions of differential accuracy across demographic groups in both DNNs and human participants. However, we also find evidence that face averages were not represented in the center of face categories. This finding should be addressed in future development of our approach, and also challenges influential cognitive models of face identity representation.
- Open Access
- 14 May 2025
- Published by Wiley in British Journal of Psychology
- Vol. 117 (2) , 656-676
The state of modelling face processing in humans with deep learning
Deep learning models trained for facial recognition now surpass the highest performing human participants. Recent evidence suggests that they also model some qualitative aspects of face processing in humans. This review compares the current understanding of deep learning models with psychological models of the face processing system. Psychological models consist of two components that operate on the information encoded when people perceive a face, which we refer to here as ‘face codes’. The first component, the core system, extracts face codes from retinal input that encode invariant and changeable properties. The second component, the extended system, links face codes to personal information about a person and their social context. Studies of face codes in existing deep learning models reveal some surprising results. For example, face codes in networks designed for identity recognition also encode expression information, which contrasts with psychological models that separate invariant and changeable properties. Deep learning can also be used to implement candidate models of the face processing system, for example to compare alternative cognitive architectures and codes that might support interchange between core and extended face processing systems. We conclude by summarizing seven key lessons from this research and outlining three open questions for future study.
- 14 May 2025
- Published by Wiley in British Journal of Psychology
- Vol. 117 (2) , 656-676
Open AccessThe state of modelling face processing in humans with deep learning
Deep learning models trained for facial recognition now surpass the highest performing human participants. Recent evidence suggests that they also model some qualitative aspects of face processing in humans. This review compares the current understanding of deep learning models with psychological models of the face processing system. Psychological models consist of two components that operate on the information encoded when people perceive a face, which we refer to here as ‘face codes’. The first component, the core system, extracts face codes from retinal input that encode invariant and changeable properties. The second component, the extended system, links face codes to personal information about a person and their social context. Studies of face codes in existing deep learning models reveal some surprising results. For example, face codes in networks designed for identity recognition also encode expression information, which contrasts with psychological models that separate invariant and changeable properties. Deep learning can also be used to implement candidate models of the face processing system, for example to compare alternative cognitive architectures and codes that might support interchange between core and extended face processing systems. We conclude by summarizing seven key lessons from this research and outlining three open questions for future study.
- Open Access
- 9 April 2025
- Published by Center for Open Science in PsyArXiv
The State of Modeling Face Processing in Humans with Deep Learning
Deep learning models trained for facial recognition now surpass the highest performing human participants. Recent evidence suggests that they also model some qualitative aspects of face processing in humans. This review compares the current understanding of deep learning models with psychological models of the face processing system. Psychological models consist of two components that operate on the information encoded when people perceive a face, which we refer to here as 'face codes.' The first component, the core system, extracts face codes from retinal input that encode invariant and changeable properties. The second component, the extended system, links face codes to personal information about a person and their social context. Studies of face codes in existing deep learning models reveal some surprising results. For example, face codes in networks designed for identity recognition also encode expression information, which contrasts with psychological models that separate invariant and changeable properties. Deep learning can also be used to implement candidate models of the face processing system, for example to compare alternative cognitive architectures and codes that might support interchange between core and extended face processing systems. We conclude by summarising seven key lessons from this research and outlining three open questions for future study.
- 9 April 2025
- Published by Center for Open Science in PsyArXiv
Open AccessThe State of Modeling Face Processing in Humans with Deep Learning
Deep learning models trained for facial recognition now surpass the highest performing human participants. Recent evidence suggests that they also model some qualitative aspects of face processing in humans. This review compares the current understanding of deep learning models with psychological models of the face processing system. Psychological models consist of two components that operate on the information encoded when people perceive a face, which we refer to here as 'face codes.' The first component, the core system, extracts face codes from retinal input that encode invariant and changeable properties. The second component, the extended system, links face codes to personal information about a person and their social context. Studies of face codes in existing deep learning models reveal some surprising results. For example, face codes in networks designed for identity recognition also encode expression information, which contrasts with psychological models that separate invariant and changeable properties. Deep learning can also be used to implement candidate models of the face processing system, for example to compare alternative cognitive architectures and codes that might support interchange between core and extended face processing systems. We conclude by summarising seven key lessons from this research and outlining three open questions for future study.
- Open Access
- 14 March 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-11
Normative face recognition ability test scores vary across online participant pools
Online participant recruitment is a cornerstone of modern psychology research. While this offers clear benefits for studying individual differences in cognitive abilities, test performance can vary across lab-based and web-based settings. Here we assess the stability of normative test scores across popular online recruitment platforms and in-person testing, for three standard measures of face identity processing ability: the GFMT2, CFMT+ , and MFMT. Participants recruited via Amazon Mechanical Turk (MTurk) scored approximately 10 percentage points lower in all tests compared to those recruited through Prolific and university students tested in the lab. Applying stricter exclusion criteria based on attention checks resulted in notably higher exclusion rates for the MTurk group (~ 62%) compared to the Prolific group (~ 22%), yet even after exclusion, some test scores remained lower for MTurk participants. Given that the GFMT2 subtests were developed using MTurk participants, we provide updated normative scores for all subtests (GFMT2-Short, GFMT2-Low, GFMT2-High) and further recommendations for their use. We also confirm the robust psychometric properties of the GFMT2-Short and GFMT2-High, demonstrating strong test–retest reliability, convergent validity with other established tests, and high diagnostic value in identifying super-recognisers. The GFMT2 subtests are freely available for use in both online and in-person research via www.gfmt2.org.
- 14 March 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-11
Open AccessNormative face recognition ability test scores vary across online participant pools
Online participant recruitment is a cornerstone of modern psychology research. While this offers clear benefits for studying individual differences in cognitive abilities, test performance can vary across lab-based and web-based settings. Here we assess the stability of normative test scores across popular online recruitment platforms and in-person testing, for three standard measures of face identity processing ability: the GFMT2, CFMT+ , and MFMT. Participants recruited via Amazon Mechanical Turk (MTurk) scored approximately 10 percentage points lower in all tests compared to those recruited through Prolific and university students tested in the lab. Applying stricter exclusion criteria based on attention checks resulted in notably higher exclusion rates for the MTurk group (~ 62%) compared to the Prolific group (~ 22%), yet even after exclusion, some test scores remained lower for MTurk participants. Given that the GFMT2 subtests were developed using MTurk participants, we provide updated normative scores for all subtests (GFMT2-Short, GFMT2-Low, GFMT2-High) and further recommendations for their use. We also confirm the robust psychometric properties of the GFMT2-Short and GFMT2-High, demonstrating strong test–retest reliability, convergent validity with other established tests, and high diagnostic value in identifying super-recognisers. The GFMT2 subtests are freely available for use in both online and in-person research via www.gfmt2.org.
- Open Access
- 1 December 2024
- Published by American Psychological Association (APA) in Journal of Experimental Psychology: Human Perception and Performance
- Vol. 50 (12) , 1143-1153
Flexible use of facial features supports face identity processing.
People prioritize diagnostic features in classification tasks. However, it is not clear whether this priority is fixed or is flexibly applied depending on the specific classification decision, or how feature use behavior contributes to individual differences in performance. Here we examined whether flexibility in features used in a face identification task supports face recognition ability. In Experiment 1, we show that the facial features most useful for identification vary-to a surprising degree-depending on the specific face identity comparison at hand. While the ears and eyes were the most diagnostic for face identification in general, they were the most diagnostic feature for just 22% and 14% of identity decisions, respectively. In three subsequent experiments, we find that flexibility in feature use contributes to an individual's face identity matching ability. Higher face identification accuracy was associated with being aware of (Experiments 2 and 4) and attending to (Experiments 3 and 4) the most diagnostic features for a specific facial comparison. This conferred an enhanced benefit relative to focusing on features that were diagnostic of face identity decisions in general (Experiment 4). We conclude that adaptability in information sampling supports face recognition ability and discuss theoretical and applied implications.
- 1 December 2024
- Published by American Psychological Association (APA) in Journal of Experimental Psychology: Human Perception and Performance
- Vol. 50 (12) , 1143-1153
Open AccessFlexible use of facial features supports face identity processing.
People prioritize diagnostic features in classification tasks. However, it is not clear whether this priority is fixed or is flexibly applied depending on the specific classification decision, or how feature use behavior contributes to individual differences in performance. Here we examined whether flexibility in features used in a face identification task supports face recognition ability. In Experiment 1, we show that the facial features most useful for identification vary-to a surprising degree-depending on the specific face identity comparison at hand. While the ears and eyes were the most diagnostic for face identification in general, they were the most diagnostic feature for just 22% and 14% of identity decisions, respectively. In three subsequent experiments, we find that flexibility in feature use contributes to an individual's face identity matching ability. Higher face identification accuracy was associated with being aware of (Experiments 2 and 4) and attending to (Experiments 3 and 4) the most diagnostic features for a specific facial comparison. This conferred an enhanced benefit relative to focusing on features that were diagnostic of face identity decisions in general (Experiment 4). We conclude that adaptability in information sampling supports face recognition ability and discuss theoretical and applied implications.