CoChat
Sign Up
Cited by 30

The DOI matches its Crossref record; 8 of the 8 references it cites were found in a registry or an index.

Diverse types of expertise in facial recognition

Towler, Alice, Dunn, James D., Castro Martínez, Sergio, Moreton, Reuben, Eklöf, Fredrick, Ruifrok, Arnout, Kemp, Richard I., White, David·2023·Scientific Reports
DOI 10.1038/s41598-023-28632-xPMC PMC10349110
Open this paper in CoChat — free account

Abstract

Facial recognition errors can jeopardize national security, criminal justice, public safety and civil rights. Here, we compare the most accurate humans and facial recognition technology in a detailed lab-based evaluation and international proficiency test for forensic scientists involving 27 forensic departments from 14 countries. We find striking cognitive and perceptual diversity between naturally skilled super-recognizers, trained forensic examiners and deep neural networks, despite them achieving equivalent accuracy. Clear differences emerged in super-recognizers’ and forensic examiners’ perceptual processing, errors, and response patterns: super-recognizers were fast, biased to respond ‘same person’ and misidentified people with extreme confidence, whereas forensic examiners were slow, unbiased and strategically avoided misidentification errors. Further, these human experts and deep neural networks disagreed on the similarity of faces, pointing to differences in their representations of faces. Our findings therefore reveal multiple types of facial recognition expertise, with each type lending itself to particular facial recognition roles in operational settings. Finally, we show that harnessing the diversity between individual experts provides a robust method of maximizing facial recognition accuracy. This can be achieved either via collaboration between experts in forensic laboratories, or most promisingly, by statistical fusion of match scores provided by different types of expert.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41598-023-28632-x

Cites

8 of 8 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Forensic facial examiners tend to be more balanced in placement of decision-making criterion between matches and mismatches and are more cautious in expressing confidence when errors are made.

    “Forensic facial examiners, on the other hand, are practitioners with a trained ability for face identification and tend to be more balanced in the placement of the decision-making criterion between matches and mismatches.These experts are also more cautious in expressing confidence when errors are made (Towler et al., 2023).”

    From A criterion-placement theory of face matching (2026) · cites this paper as Towler et al., 2023

  • Super-recognizers exhibit a tendency to classify face pairs as matches and are biased to make highly confident errors on mismatch trials.

    “These observers exhibit a tendency to classify face pairs as matches and are biased to make highly confident errors on mismatch trials (Towler et al., 2023).”

    From A criterion-placement theory of face matching (2026) · cites this paper as Towler et al., 2023

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 73

  1. Bruce, V., Henderson, Z., Newman, C. & Burton, A. M. Matching identities of familiar and unfamiliar faces caught on CCTV images. J. Exp. Psychol. Appl. 7, 207–218 (2001). · DOI · Paper page
  2. White, D., Kemp, R. I., Jenkins, R., Matheson, M. & Burton, A. M. Passport officers’ errors in face matching. PLoS ONE 9, 1–6. https://doi.org/10.1371/journal.pone.0103510 (2014). · DOI · Paper page
  3. Wirth, B. E. & Carbon, C. C. An easy game for frauds? Effects of professional experience and time pressure on passport-matching performance. J. Exp. Psychol. Appl. 23, 138–157. https://doi.org/10.1037/xap0000114 (2017). · DOI · Paper page
  4. White, D., Towler, A. & Kemp, R. I. 2021 Forensic face matching: Research and practice M. Bindemann (ed.), Oxford University Press, Oxford.
  5. Moreton, R. In Forensic face matching: Research and practice (ed M. Bindemann) (Oxford University Press, 2021). · DOI
  6. Wilkinson, C. & Evans, R. Are facial image analysis experts any better than the general public at identifying individuals from CCTV images?. Sci. Justice 49, 191–196 (2009). · DOI · Paper page
  7. Norell, K. et al. The effect of image quality and forensic expertise in facial image comparisons. J. Forensic Sci. 60, 331–340. https://doi.org/10.1111/1556-4029.12660 (2015). · DOI · Paper page
  8. Albright, T. D. The US department of justice stumbles on visual perception. Proc. Natl. Acad. Sci. 118, e2102702118. https://doi.org/10.1073/pnas.2102702118 (2021). · DOI
  9. Bell, S. et al. A call for more science in forensic science. Proc. Natl. Acad. Sci. https://doi.org/10.1073/pnas.1712161115 (2018). · DOI
  10. Edmond, G. et al. Model forensic science. Aust. J. Forensic Sci. 57, 144–154. https://doi.org/10.1016/j.scijus.2016.11.005 (2016). · DOI · Paper page
  11. Towler, A., Kemp, R. I. & White, D. In Forensic face matching: Research and practice (ed M. Bindemann) (Oxford University Press, 2021).
  12. Russell, R., Duchaine, B. & Nakayama, K. Super-recognizers: People with extraordinary face recognition ability. Psychon. Bull. Rev. 16, 252–257. https://doi.org/10.3758/PBR.16.2.252 (2009). · DOI · Paper page
  13. Shakeshaft, N. G. & Plomin, R. Genetic specificity of face recognition. Proc. Natl. Acad. Sci. 112, 12887–12892. https://doi.org/10.1073/pnas.1421881112 (2015). · DOI
  14. Wilmer, J. B. et al. Human face recognition ability is specific and highly heritable. Proc. Natl. Acad. Sci. 107, 5238–5241. https://doi.org/10.1073/pnas.0913053107 (2010). · DOI · Paper page
  15. Dunn, J. D., Towler, A., Kemp, R. I. & White, D. Selecting police super-recognisers. PLoS ONE 18(5), e0283682. https://doi.org/10.1371/journal.pone.0283682 (2023). · DOI
  16. Ramon, M., Bobak, A. K. & White, D. Super-recognizers: From the lab to the world and back again. Br. J. Psychol. 110, 461–479. https://doi.org/10.1111/bjop.12368 (2019). · DOI
  17. Robertson, D. J., Noyes, E., Dowsett, A. J., Jenkins, R. & Burton, A. M. Face recognition by metropolitan police super-recognisers. PLoS ONE https://doi.org/10.1371/journal.pone.0150036 (2016). · DOI
  18. Phillips, P. J. et al. Face recognition accuracy in forensic examiners, super-recognisers and algorithms. Proc. Natl. Acad. Sci. https://doi.org/10.1073/pnas.1721355115 (2018). · DOI · Paper page
  19. Grother, P. & Ngan, M. Face recognition vendor test (FRVT). (National Institute of Standards and Technology, 2014). · DOI · Paper page
  20. Grother, P., Ngan, M. & Hanaoka, K. Ongoing face recognition vendor test (FRVT) part 1: Verification. National Institute of Standards and Technology (2018). · DOI
  21. Grother, P., Ngan, M. & Hanaoka, K. Ongoing face recognition vendor test (FRVT) part 1: Verification. National Institute of Standards and Technology (2019). · DOI
  22. Blauch, N. M., Behrmann, M. & Plaut, D. C. Computational insights into human perceptual expertise for familiar and unfamiliar face recognition. Cognition 208, 104341. https://doi.org/10.1016/j.cognition.2020.104341 (2021). · DOI
  23. Jacob, G., Pramod, R. T., Katti, H. & Arun, S. P. Qualitative similarities and differences in visual object representations between brains and deep networks. Nat. Commun. 12, 1872. https://doi.org/10.1038/s41467-021-22078-3 (2021). · DOI
  24. O’Toole, A. J., Castillo, C. D., Parde, C. J., Hill, M. Q. & Chellappa, R. Face Space Representations in Deep Convolutional Neural Networks. Trends Cogn. Sci. 22, 794–809. https://doi.org/10.1016/j.tics.2018.06.006 (2018). · DOI
  25. Jozwik, K. M., O’Keeffe, J., Storrs, K. R. & Kriegeskorte, N. Face dissimilarity judgements are predicted by representational distance in deep neural networks and principal-component face space. Biorxiv https://doi.org/10.1101/2021.04.09.438859 (2021). · DOI
  26. Hancock, P. J., Somai, R. S., Mileva, V. R. & Bobak, A. K. Convolutional neural net face recognition works in non-human-like ways. R. Soc. Open Sci. 7, 200595 (2020). · DOI
  27. Grossman, S. et al. Convergent evolution of face spaces across human face-selective neuronal groups and deep convolutional networks. Nat. Commun. 10, 4934. https://doi.org/10.1038/s41467-019-12623-6 (2019). · DOI
  28. Geisler, W. S. Contributions of ideal observer theory to vision research. Vision. Res. 51, 771–781. https://doi.org/10.1016/j.visres.2010.09.027 (2011). · DOI
  29. Burton, A. M., White, D. & McNeill, A. The Glasgow Face Matching Test. Behav. Res. Methods 42, 286–291. https://doi.org/10.3758/BRM.42.1.286 (2010). · DOI · Paper page
  30. Growns, B., Dunn, J. D., Mattijssen, E. J. A. T., Quigley-McBride, A. & Towler, A. Match me if you can: Evidence for a domain-general visual comparison ability. Psychon. Bull. Rev. https://doi.org/10.3758/s13423-021-02044-2 (2022). · DOI
  31. White, D., Phillips, P. J., Hahn, C. A., Hill, M. & O’Toole, A. J. Perceptual expertise in forensic facial image comparison. Proc. R. Soc. London B: Biol. Sci. 282, 1814–1822. https://doi.org/10.1098/rspb.2015.1292 (2015). · DOI · Paper page
  32. Towler, A., White, D. & Kemp, R. I. Evaluating the feature comparison strategy for forensic face identification. J. Exp. Psychol. Appl. 23, 47–58. https://doi.org/10.1037/xap0000108 (2017). · DOI · Paper page
  33. Rice, A., Phillips, P. J., Natu, V., An, X. & O’Toole, A. J. Unaware person recognition from the body when face identification fails. Psychol. Sci. 24, 2235–2243 (2013). · DOI
  34. White, D., Burton, A. M., Kemp, R. I. & Jenkins, R. Crowd effects in unfamiliar face matching. Appl. Cogn. Psychol. 27, 769–777 (2013). · DOI · Paper page
  35. Kittler, J., Hatef, M., Duin, R. P. W. & Matas, J. On combining classifiers. IEEE Trans. Pattern Anal. Mach. Intell. 20, 226–239 (1998). · DOI
  36. Jeckeln, G., Hahn, C. A., Noyes, E., Cavazos, J. G. & O’Toole, A. J. Wisdom of the social versus non-social crowd in face identification. Br. J. Psychol. 109, 724–735. https://doi.org/10.1111/bjop.12291 (2018). · DOI · Paper page
  37. Dunn, J. D. et al. Face-Information Sampling in Super-Recognizers. Psychol. Sci. 33, 1615–1630. https://doi.org/10.1177/09567976221096320 (2022). · DOI
  38. Kahneman, D. & Klein, G. Conditions for intuitive expertise: A failure to disagree. Am. Psychol. 64, 515–526. https://doi.org/10.1037/a0016755 (2009). · DOI · Paper page
  39. Edmond, G. & Wortley, N. Interpreting image evidence: Facial mapping, police familiars and super-recognisers in England and Australia. J. Int. Comp. Law 3, 473–522 (2016).
  40. Edmond, G., White, D., Towler, A., San Roque, M. & Kemp, R. I. Facial recognition and image comparison evidence: Identification by investigators, familiars, experts, super-recognisers and algorithms. Melbourne University Law Review (in press).
  41. Davis, J. P., Lander, K., Evans, R. & Jansari, A. Investigating predictors of superior face recognition ability in police super-recognisers. Appl. Cogn. Psychol. 30, 827–840 (2016). · DOI
  42. Davis, J. P., Forrest, C., Tremi, F. & Jansari, A. Identification from CCTV: Assessing police super-recogniser ability to spot faces in a crowd and susceptibility to change. Appl. Cogn. Psychol. https://doi.org/10.1002/acp.3405 (2018). · DOI
  43. Hill, M. Q. et al. Deep convolutional neural networks in the face of caricature. Nat. Mach. Intell. 1, 522–529. https://doi.org/10.1038/s42256-019-0111-7 (2019). · DOI
  44. Vogelsang, L. et al. Potential downside of high initial visual acuity. Proc. Natl. Acad. Sci. 115, 11333. https://doi.org/10.1073/pnas.1800901115 (2018). · DOI · Paper page
  45. Zhuang, C. et al. Unsupervised neural network models of the ventral visual stream. Proc. Natl. Acad. Sci. 118, e2014196118. https://doi.org/10.1073/pnas.2014196118 (2021). · DOI
  46. Taylor, J. E. T. & Taylor, G. W. Artificial cognition: How experimental psychology can help generate explainable artificial intelligence. Psychon. Bull. Rev. 28, 454–475. https://doi.org/10.3758/s13423-020-01825-5 (2021). · DOI
  47. Lin, Y.-S. et al. xCos: An explainable cosine metric for face verification task. arxiv (2021). · DOI
  48. Fussey, P., Davies, B. & Innes, M. ‘Assisted’ facial recognition and the reinvention of suspicion and discretion in digital policing. Br. J. Criminol. 61, 325–344. https://doi.org/10.1093/bjc/azaa068 (2020). · DOI
  49. Dror, I. E. & Langenburg, G. “Cannot decide”: The fine line between appropriate inconclusive determinations versus unjustifiably deciding not to decide. J. Forensic Sci. 64, 10–15. https://doi.org/10.1111/1556-4029.13854 (2019). · DOI
  50. Towler, A. et al. Are forensic scientists experts?. J. Appl. Res. Mem. Cogn. 7, 199–208 (2018). · DOI · Paper page
  51. Del Grecco, K. J. Facial recognition technology: Ensuring transparency in government use (Federal Bureau of Investigation, 2019).
  52. Diamond, R. & Carey, S. Why faces are and are not special: An effect of expertise. J. Exp. Psychol. Gen. 115, 107–117 (1986). · DOI
  53. Wang, R., Li, J., Fang, H., Tian, M. & Liu, J. Individual differences in holistic processing predict face recognition ability. Psychol. Sci. 23, 169–177. https://doi.org/10.1177/0956797611420575 (2012). · DOI
  54. Yildirim, I., Belledonne, M., Freiwald, W. & Tenenbaum, J. Efficient inverse graphics in biological face processing. Sci. Adv. 6, eaax5979 (2020). · DOI
  55. Dowsett, A. J. & Burton, A. M. Unfamiliar face matching: Pairs out-perform individuals and provide a route to training. Br. J. Psychol. 106, 433–445. https://doi.org/10.1111/bjop.12103 (2014). · DOI · Paper page
  56. Bobak, A. K., Pampoulov, P. & Bate, S. Detecting superior face recognition skills in a large sample of young British adults. Front. Psychol. 7, 1378 (2016). · DOI
  57. McKone, E. et al. Face ethnicity and measurement reliability affect face recognition performance in developmental prosopagnosia: Evidence from the Cambridge face memory test-Australian. Cogn. Neuropsychol. 28, 109–146 (2011). · DOI · Paper page
  58. Dunn, J. D., Summersby, S., Towler, A., Davis, J. P. & White, D. UNSW Face Test: A screening test for super-recognisers. PLoS ONE 15, e0241747 (2020). · DOI
  59. Jenkins, R., White, D., Van Montfort, X. & Burton, A. M. Variability in photos of the same face. Cognition 121, 313–323. https://doi.org/10.1016/j.cognition.2011.08.001 (2011). · DOI
  60. Tangen, J. M., Thompson, M. B. & McCarthy, D. J. Identifying fingerprint expertise. Psychol. Sci. 22, 995–997. https://doi.org/10.1177/0956797611414729 (2011). · DOI
  61. Kagan, J. Reflection-impulsivity and reading ability in primary grade children. Child Dev. 36, 609–628 (1965). · DOI
  62. Duchaine, B. C. & Nakayama, K. The Cambridge Face Memory Test: Results for neurologically intact individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants. Neuropsychologia 44, 576–585. https://doi.org/10.1016/j.neuropsychologia.2005.07.001 (2006). · DOI
  63. Ballantyne, K., Edmond, G. & Found, B. Peer review in forensic science. Forensic Sci. Int. 277, 66–76 (2017). · DOI
  64. Simonyan, K. & Zisserman, A. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014). · arXiv
  65. He, K., Zhang, X., Ren, S. & Sun, J. In 2016 IEEE conference on computer vision and pattern recognition (CVPR) 770–778 (2016).
  66. Ng, H. & Winkler, S. In 2014 IEEE international conference on image processing (ICIP). 343–347. · DOI
  67. King, D. E. Dlib-ml: A machine learning toolkit. J. Mach. Learn. Res. 10, 1755–1758 (2009).
  68. Cao, Q., Shen, L., Xie, W., Parkhi, O. M. & Zisserman, A. In 2018 13th IEEE international conference on automatic face & gesture recognition (FG 2018) 67–74 (2018).
  69. Guo, Y., Zhang, L., Hu, Y., He, X. & Gao, J. Computer vision—ECCV 2016 87–102 (Springer International Publishing, 2016).
  70. Hu, J., Shen, L. & Sun, G. In Proceedings of the IEEE conference on computer vision and pattern recognition 7132–7141 (2018).
  71. Yi, D., Lei, Z., Liao, S. & Li., S. Z. Learning face representation from scratch. arXiv 1411.7923 (2014). · arXiv
  72. Serengil, S. I. & Ozpinar, A. In 2020 IEEE innovations in intelligent systems and applications conference. 23–27.
  73. Schroff, F., Kalenichenko, D. & Philbin, J. Facenet: A unified embedding for face recognition and clustering. In Proceedings of the IEEE conference on computer vision and pattern recognition, 815–823 (2015). · DOI · Paper page

Cited by

Sorted by citations

Metadata

Authors
Towler, A., Dunn, J. D., Castro Martínez, S., Moreton, R., Eklöf, F., Ruifrok, A., Kemp, R. I., & White, D.
Venue
Scientific Reports
Publisher
Springer Science and Business Media LLC
Published
2023-07-14
Volume
13
Issue
1
ISSN
2045-2322
DOI
10.1038/s41598-023-28632-x
PMC
PMC10349110
Cited by
30
References
8
Language
en
Licence
Funders
  • ARC Linkage Project (LP160101523)
  • ARC DECRA (DE210100357)
  • ARC Discovery Project (DP190100957)

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review