Exploring Spatial Frequency and Orientation Tunings for Face Recognition in Eight Cultural Groups

Poster Presentation 63.428: Wednesday, May 22, 2024, 8:30 am – 12:30 pm, Pavilion
Session: Face and Body Perception: Disorders, individual differences

Francis Gingras1,2 (), Arianne Richer2, Cousineau Alex2, Justin Duncan2, Daniel Fiset2, Frédéric Gosselin3, Caroline Blais2; 1Université du Québec à Montréal, 2Université du Québec en Outaouais, 3Université de Montréal

East Asians use lower spatial frequencies (SF) compared to Westerners while processing faces (Tardif et al., 2017). These differences have been attributed to culture; however, the underlying mechanism remains unclear. While many hypotheses exist (e.g. social orientation, urbanisation), having data for only two cultural groups makes generalisation difficult/iffy at best. The present study addresses this limitation by measuring SF tunings across eight cultural groups. Preliminary data was collected in Sub-Saharan Africa (n = 70), East Asia (n = 45), Eastern Europe (n = 83), English speaking countries (n = 63), Latin America (n = 89), Middle East (n=50), Southern Asia (n=72) and Western Europe (n=78). Targeted sample size is n=80 for all groups, as pre-registered on the OSF. Participants completed 600 trials of a same/different face matching task online using VPixx Pack & Go (VPixx Technologies, 2021). Target stimuli were filtered using SFO Bubbles, allowing for the sampling of all combinations of SF and orientations (Gingras et al., 2022). A weighted sum of all filters was computed to reveal SFO use for each participant as a 2D classification image. Preliminary analyses comparing the top 1% of t-scores across cultures reveal no differences in orientation tunings but reveal that Eastern and Southern Asians, as well as Sub-Saharan Africans, use lower SF compared to Western Europe/English countries. This is inconsistent with the recently proposed urbanization hypothesis (Caparos et al., 2012), according to which African cultures should show a local bias (and therefore use higher spatial frequencies). While the social orientation hypothesis is more consistent with our results, it fails to predict other visual effects, such as the Ebbinghaus illusion (Caparos et al., 2012) or eye movements (Gingras et al., in press). Other theories, applicable not only to East Asia, but to Southern Asia and Sub-Saharan Africa as well, should be explored.