Purchase this article with an account.
Shabnam Rostamirad, Ipek Oruc, Jason J. S. Barton; Face space has a center-surround organization: Evidence from a novel contrast-based face-adaptation technique. Journal of Vision 2008;8(6):1144. doi: https://doi.org/10.1167/8.6.1144.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Face-space' is a concept of the multidimensional representation of faces in the neural system. Faces of similar appearance are considered closer to each other than faces that are dissimilar; however, it is not clear how representations interact across this space. In this study we used a novel adaptation paradigm in which we examined how a 200ms preview of one face affected contrast thresholds for recognition of the same face or of other faces of the same or different ethnicity as the previewed face. We obtained images of two Caucasian and two Asian female faces. We used an ideal observer analysis and measured pair-wise discrimination thresholds to verify that, in this stimulus set, faces were more similar to those of the same ethnic group than to those of the other. In the adaptation experiment, 7 subjects viewed either a blank screen or one of the four faces for 200ms, followed by a 150ms test face, which was a low-contrast version of one of the same four faces. Subjects indicated which of the four faces they believed the test face to have been. Test-face contrast was varied with the Quest procedure to estimate 82% contrast thresholds. We found that a 200ms preview of a face facilitated its recognition but impaired recognition of other faces. Thresholds were more elevated for other faces of the same ethnic group than for other faces of a different ethnic group. This is consistent with a center-surround organization in which the neural representations of faces that are closer in face-space to the priming stimulus are suppressed more than those more distant. This suggests that center-surround relationships that serve to sharpen neural tuning and perceptual discrimination are also found in high-level face representations.
This PDF is available to Subscribers Only