The performance of practitioners conducting facial comparisons on images of children across age


Autoři: Dana Michalski aff001;  Rebecca Heyer aff001;  Carolyn Semmler aff002
Působiště autorů: Defence Science and Technology Group, Edinburgh, South Australia, Australia aff001;  School of Psychology, University of Adelaide, Adelaide, South Australia, Australia aff002
Vyšlo v časopise: PLoS ONE 14(11)
Kategorie: Research Article
doi: 10.1371/journal.pone.0225298

Souhrn

Determining the identity of children is critical to aid in the fight against child exploitation, as well as for passport control and visa issuance purposes. Facial image comparison is one method that may be used to determine identity. Due to the substantial amount of facial growth that occurs in childhood, it is critical to understand facial image comparison performance across both chronological age (the age of the child), and age variation (the age difference between images). In this study we examined the performance of 120 facial comparison practitioners from a government agency on a dataset of 23,760 image pairs selected from the agency’s own database of controlled, operational images. Each chronological age in childhood (0–17 years) and age variations ranging from 0–10 years were examined. Practitioner performance was found to vary considerably across childhood, and depended on whether the pairs were mated (same child) or non-mated (different child). Overall, practitioners were more accurate and confident with image pairs containing older children, and also more accurate and confident with smaller age variations. Chronological age impacted on accuracy with mated pairs, but age variation did not. In contrast, both age and age variation impacted on accuracy with non-mated pairs. These differences in performance show that changes in the face throughout childhood have a significant impact on practitioner performance. We propose that improvements in accuracy may be achievable with a better understanding of which facial features are most appropriate to compare across childhood, and adjusting training and development programs accordingly.

Klíčová slova:

Algorithms – Face – Face recognition – Children – Imaging techniques – Infants – Nose – Sense of agency


Zdroje

1. Prince J. To examine emerging police use of facial recognition systems and facial image comparison procedures. 2012; https://www.churchilltrust.com.au/media/fellows/2012_Prince_Jason.pdf.

2. Michalski D. The impact of age-related variables on facial comparisons with images of children: Algorithm and practitioner performance [Doctoral Dissertation]: University of Adelaide; 2017.

3. Megreya AM, Burton AM. Hits and false positives in face matching: A familiarity-based dissociation. Perception & psychophysics. 2007;69(7):1175–84.

4. Heyer R, MacLeod V, Carter L, Semmler C, Ma-Wyatt A. Profiling the facial comparison practitioner in Australia. DST Edinburgh, South Australia: DST-Group-GD-1030; 2017.

5. Towler A, White D, Kemp RI. Evaluating the feature comparison strategy for forensic face identification. Journal of Experimental Psychology: Applied. 2017;23(1):47. doi: 10.1037/xap0000108 28045276

6. FISWG. Recommendations for a training program in facial comparison. Facial Identification Scientific Working Group; 2012. p. 1–6.

7. Phillips PJ, Yates AN, Hu Y, Hahn CA, Noyes E, Jackson K, et al. Face recognition accuracy of forensic examiners, superrecognizers, and face recognition algorithms. Proceedings of the National Academy of Sciences. 2018:201721355.

8. White D, Kemp RI, Jenkins R, Matheson M, Burton AM. Passport officers’ errors in face matching. PloS one. 2014;9(8):e103510. doi: 10.1371/journal.pone.0103510 25133682

9. White D, Phillips PJ, Hahn CA, Hill M, O'Toole AJ. Perceptual expertise in forensic facial image comparison. Proc Biol Sci. 2015;282(1814). Epub 2015/09/04. doi: 10.1098/rspb.2015.1292 26336174; PubMed Central PMCID: PMC4571699.

10. Heyer R, Semmler C, Hendrickson AT. Humans and algorithms for facial recognition: The effects of candidate list length and experience on performance. Journal of applied research in memory and cognition. 2018;7(4):597–609.

11. Megreya AM, Sandford A, Burton AM. Matching face images taken on the same day or months apart: The limitations of photo ID. Applied Cognitive Psychology. 2013;27(6):700–6.

12. Michalski D, Yiu SY, Malec C. The impact of age and threshold variation on facial recognition algorithm performance using images of children. International Conference on Biometrics (ICB): IEEE; 2018. p. 217–24.

13. White D, Dunn JD, Schmid AC, Kemp RI. Error rates in users of automatic face recognition software. PLoS One. 2015;10(10):e0139827. doi: 10.1371/journal.pone.0139827 26465631

14. Kramer RS, Mulgrew J, Reynolds MG. Unfamiliar face matching with photographs of infants and children. PeerJ. 2018;6:e5010. doi: 10.7717/peerj.5010 29910991

15. Ferguson EL. Facial identification of children: a test of automated facial recognition and manual facial comparison techniques on juvenile face images [Doctoral dissertation]: University of Dundee; 2015.

16. Zhang S. Smuggling and trafficking in human beings: all roads lead to America. Westport, CT: Praeger Publishers; 2007.

17. Hole M, McLindin B, Hanton K, Malec C, Yiu SY, Hanly G. An overview of a DSTO developed human operator image comparison software tool—Comparer. Edinburgh, South Australia: DSTO-GD-0855; 2015.

18. Kozak FK, Ospina JC, Cardenas MF. Characteristics of normal and abnormal postnatal craniofacial growth and development. 2015. In: Cummings Pediatric Otolaryngology E-Book [Internet]. [55–80].

19. Ricanek K, Mahalingam G, Albert AM, Bruegge RV, Fairhurst M. Human face ageing: a perspective analysis from anthropometry and biometrics. Book Chapter in Age Factors in Biometric Processing Edited by Michael Fairhurst. 2013.

20. Towler A, White D, Kemp RI. Evaluating training methods for facial image comparison: The face shape strategy does not work. Perception. 2014;43(2–3):214–8. doi: 10.1068/p7676 24919354

21. Grother P, Ngan M. Face Recognition Vendor Test (FRVT)—Performance of face identification algorithms. NIST Interagency Report 8009; 2014.

22. Wilkinson C. Juvenile facial reconstruction. In: Wilkinson CR C, editor. Forensic facial reconstruction. Cambridge, New York: Cambridge University Press; 2012. p. 254–60.

23. Vernon MD. A further study of visual perception: Cambridge University Press; 1952.

24. Bruce V. Stability from variation: The case of face recognition the MD Vernon memorial lecture. The Quarterly Journal of Experimental Psychology. 1994;47(1):5–28. doi: 10.1080/14640749408401141 8177963

25. Stephens RG, Semmler C, Sauer JD. The effect of the proportion of mismatching trials and task orientation on the confidence–accuracy relationship in unfamiliar face matching. Journal of Experimental Psychology: Applied. 2017;23(3):336. doi: 10.1037/xap0000130 28805443

26. Havard C. Eye movement strategies during face matching: University of Glasgow; 2007.

27. Megreya AM. Feature-by-feature comparison and holistic processing in unfamiliar face matching. PeerJ. 2018;6:e4437. doi: 10.7717/peerj.4437 29503772

28. FISWG. Guidelines for facial comparison methods. Facial Identification Scientific Working Group; 2012. p. 1–15.

29. FISWG. Physical stability of facial features of adults. Facial Identification Scientific Working Group; 2019. p. 1–16.

30. Farkas LG, Hreczko TA. Age-related changes in selected linear and angular measurements of the craniofacial complex in healthy North American Caucasians. In: Farkas LG, editor. Anthropometry of the Head and Face. second ed ed. New York: Raven Press, Ltd.; 1994. p. 89–102.

31. Yadav D, Singh R, Vatsa M, Noore A. Recognizing age-separated face images: Humans and machines. PloS one. 2014;9(12):e112234. doi: 10.1371/journal.pone.0112234 25474200


Článek vyšel v časopise

PLOS One


2019 Číslo 11