Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

A novel interactive face matching procedure: Performance of normal and super face recognizers

142 views

Published on

A new procedure to increase face matching accuracy in forensic face examination. By Harriet M. J. Smith , Sally Andrews , David White , Josh P. Davis , Melissa F. Colloff, Thom S. Baguley , & Heather D. Flowe

Published in: Education
  • Be the first to comment

  • Be the first to like this

A novel interactive face matching procedure: Performance of normal and super face recognizers

  1. 1. A novel interactive face matching procedure: Performance of normal and super face recognizers Harriet M. J. Smith , Sally Andrews , David White , Josh P. Davis , Melissa F. Colloff, Thom S. Baguley , & Heather D. Flowe
  2. 2. Face matching 11 October 2018 2 Are we really that good at recognising/perceiving faces? Unfamiliar face matching is error prone (Bruce et al., 1999) Passport officers – 10% errors (White et al., 2014) ?
  3. 3. Why is face matching difficult? Between-person and within-person variability Unfamiliar face matching accuracy improves when information about variability is available (Menon et al., 2015) 11 October 2018 3
  4. 4. Orientation information 11 October 2018 4 Faces look different from different viewpoints Recognition memory – differences in viewpoint affect accuracy (Bruce, 1982) Less of an effect for face matching? (Estudillo & Bindemann, 2014 but see Bruce et al., 1999; Hill & Bruce, 1996) No benefit when both frontal and profile views were provided (Kramer & Reynolds, 2018) 3D view-independent representation? Images from UNSW Unfamiliar Face and Voice Database (White, Burton & Kemp, 2016)
  5. 5. Interactivity What if the participant could interact with the face, and manoeuvre it into any viewpoint? - Facilitate building of 3D view independent representation? - Encourage engagement  deep encoding? 11 October 2018 5 ?
  6. 6. Normal vs super recognisers 11 October 2018 6 Individual differences in unfamiliar face recognition and matching (e.g. Davis et al., 2016; Bobak et al., 2015; Russell et al., 2009) Super-recognisers – natural face recognition/face perception skills What are SRs doing differently? (see Bobak et al., 2015) - Holistic information? - Structural encoding  view-independent representation?
  7. 7. Aims How accurate? Effect of orientation information/movement/interactivity? Accuracy/confidence relationship? Experiment 1: ‘Normal’ face recognisers Experiment 2: ‘Superior’ face recognisers 11 October 2018 7
  8. 8. Experiment 1 ‘Normal’ face recognisers
  9. 9. Design 4 x 2 mixed factorial design Within-subjects factor: identity (same or different) Between subjects factor: to-be-compared image (frontal, orientations, moving, interactive) 11 October 2018 9 Face 1 Face 2
  10. 10. Participants N = 306 176 female, 130 male Age range 18-67 (M = 33.47, SD = 10.71) CFMT+ ≤ 92 (M = 76.34, SD = 9.32) 11 October 2018 10
  11. 11. Stimuli 11 October 2018 11 UNSW Unfamiliar Face and Voice Database (White, Burton & Kemp, 2016)
  12. 12. 11 October 2018 12 Frontal
  13. 13. Orientations 11 October 2018 13
  14. 14. Moving 11 October 2018 14
  15. 15. Interactive 11 October 2018 15
  16. 16. Results: Accuracy 11 October 2018 16
  17. 17. Results: Accuracy (MLM) No main effect of identity (p = .085) Main effect of image type (p = .019) Interaction: image type*identity (p < .001) 11 October 2018 17
  18. 18. Results: Confidence 11 October 2018 18
  19. 19. Results: Confidence and accuracy 11 October 2018 19 Frontal Orientations Moving Interactive
  20. 20. Experiment 2 ‘Superior’ face recognisers
  21. 21. Participants N = 57 28 female, 29 male Age range 18 - 70 (M = 34.91, SD = 10.56) CFMT+ ≥ 93 (M = 95.19, SD = 1.76) 11 October 2018 21 (‘superior’ not SR)
  22. 22. Results: Accuracy 11 October 2018 22
  23. 23. Results: Accuracy (MLM) Main effect of identity (p = .011) No main effect of image type (p = .222) Interaction: identity*image condition (p < .001) 11 October 2018 23
  24. 24. Results: Confidence 11 October 2018 24
  25. 25. Results: Confidence and accuracy 11 October 2018 25 Frontal Orientations Moving Interactive
  26. 26. Conclusions 11 October 2018 26 Normal and superior face recognisers: different patterns of performance Interactive makes ‘normal’ face recognisers behave like ‘superior’ face recognisers when viewing static images Normal Superior
  27. 27. Should we recommend interactivity in an applied setting? 11 October 2018 27 Conclusions
  28. 28. Thank you for listening @harrietsmith15 Harriet M. J. Smith harriet.smith02@ntu.ac.uk

×