Purchase this article with an account.
Matthew V Pachai, Mitchel Downham, Jennifer K E Steeves; Horizontal selectivity during face perception in the visual periphery. Journal of Vision 2019;19(10):228c. doi: https://doi.org/10.1167/19.10.228c.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Several reports have demonstrated selective processing of horizontally-oriented spatial frequency components during face-related tasks (e.g. Goffaux & Dakin, Front Psychol, 2010; Pachai et al, Front Psychol, 2013), but these studies were all conducted using foveal stimuli, and it remains unclear whether the same patterns will hold in the periphery. We know that diagnostic information is rapidly extracted from peripheral faces, as faces are detected faster than other categories of objects at 8° eccentricity (Crouzet et al, J Vis, 2010), and observers typically direct their first saccade to an optimal landing point on a face (Peterson & Eckstein, PNAS, 2012). Therefore, to investigate whether peripheral face processing is also horizontally-selective, we had observers complete a 4-AFC face discrimination task with targets presented either in the fovea or at 8° eccentricity, filtered to retain either horizontal or vertical frequency components. We observed no significant interaction between eccentricity and filter orientation in this task, suggesting that horizontal selectivity is deployed in the periphery as it is in the fovea. In a second experiment, we investigated whether face crowding is horizontally selective. Specifically, we had observers complete a same-different discrimination task with one face presented at the fovea and one presented at 8° eccentricity. On every trial, the peripheral face was presented alone or flanked by six different faces. Across trials, these faces were filtered to retain horizontal, vertical, or all frequency components. We observed lower performance when the peripheral face was flanked, but no difference in the magnitude of crowding exerted by horizontal, vertical, or unfiltered flankers. Together, these results suggest that horizontal selectivity is deployed in the periphery, but that the mere presence of upright flankers is sufficient to induce face crowding. We are conducting follow-up studies to explore the implications of this result.
This PDF is available to Subscribers Only