September 2018
Volume 18, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2018
Task-specific extraction of horizontal information in faces
Author Affiliations
  • Gabrielle Dugas
    Université du Québec en Outaouais
  • Jessica Royer
    Université du Québec en Outaouais
  • Justin Duncan
    Université du Québec en OutaouaisUniversité du Québec à Montréal
  • Caroline Blais
    Université du Québec en Outaouais
  • Daniel Fiset
    Université du Québec en Outaouais
Journal of Vision September 2018, Vol.18, 931. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Gabrielle Dugas, Jessica Royer, Justin Duncan, Caroline Blais, Daniel Fiset; Task-specific extraction of horizontal information in faces. Journal of Vision 2018;18(10):931.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Horizontal information is crucial for accurate face processing (Goffaux & Dakin, 2010). Individual differences in horizontal tuning were shown to correlate with aptitude levels in both face identification (Pachai, Sekuler & Bennett, 2013) and facial expression categorization (Duncan et al., 2017).These results thus indicate that the same visual information correlates with abilities in two different face processing tasks. Here, we intended to verify if the ability to extract horizontal information generalizes from one task to the other at the individual level. To do this, we asked 28 participants to complete both a 10-AFC face identification task and a race categorization (Caucasian vs. African-American) task (600 trials per task). To find out which parts of the orientation spectrum were associated with accuracy, images were randomly filtered with orientation bubbles (Duncan et al., 2017). We then performed, for each subject, what amounts to a multiple linear regression of orientation sampling vectors (independent variable) on response accuracy scores (dependent variable). A group classification vector (CV) was created by first summing individually z-scored CVs across subjects, and then dividing the outcome by √n, where n is the sample size. These analyses, performed separately for each task, show that horizontal information is highly diagnostic for both face identification (Zmax = 24.8) and race categorization (Zmax = 22.9), all ps < .05 and Group CVs of both task were highly correlated, r= .96, p< .001, showing high similarity in visual strategies at the group level. At the individual level, however, horizontal tuning measures (as per Duncan et al., 2017) in the identification and race categorization tasks did not correlate, r = -0.02, ns. Our results thus show that, although horizontal information is diagnostic for both tasks, individual differences in the extraction of this information appears to be task dependent.

Meeting abstract presented at VSS 2018


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.