Purchase this article with an account.
Logan Trujillo, Erin Anderson, Judith Langlois; Visual Perception of Facial Attractiveness and Typicality Reflects an Ideal Dimension of Face Category Structure. Journal of Vision 2018;18(10):1335. doi: 10.1167/18.10.1335.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Previous evidence suggests a relationship between the visual perception of human facial attractiveness and the perceived typicality (degree of category representativeness) of a face, with typical faces perceived as attractive and atypical faces perceived as unattractive. However, the relationship between facial attractiveness and the psychological representation of typicality remains unclear. We examined this relationship by gathering human judgments of the attractiveness, typicality, and pairwise similarity of 100 young adult Caucasian female faces. We then described the perceived attractiveness and typicality judgments by fitting three computational models of facial attractiveness and typicality as mathematically expressed within a perceptual "face space" derived from the similarity ratings via multidimensional scaling: 1) the Generalized Context Model, in which attractiveness/typicality is a function of the similarity of a face to other face exemplars in a population of faces; 2) the Central Prototype Model, in which attractiveness/typicality is a function of the similarity of a face to the central tendency of a population of faces; and 3) the Ideal Dimension Model, in which attractiveness/typicality is a function of the distance of a face along a category dimension that reflects a combination of ideal face features. The Ideal Dimension Model described attractiveness and typicality better than the Generalized Context Model or the Central Prototype Model as quantified via measures of model goodness of fit and model generalizability. Our findings suggest that facial attractiveness and typicality are best captured by a single ideal dimension of face category structure.
Meeting abstract presented at VSS 2018
This PDF is available to Subscribers Only