Abstract
Purpose: Adaptation to female faces makes a face that was previously seen as gender-neutral appear male (Webster et al., 2004). This face adaptation technique has been widely used as a tool for examining selectivity for properties such as identity and expression. Here, we asked what might be the perceptual benefit of selective face adaptation by examining whether adaptation affects the relative salience of faces. Methods: Subjects were pre-adapted for 3 minutes to a series of faces (1 face/sec). Adaptor faces belonged to 3 of 4 categories, for example: male Asian, female Caucasian, and male Caucasian. On each trial subjects were asked to detect target faces in either a rapid serial visual presentation (RSVP) or visual search (VS) task. Each subject was tested on two target categories: an unadapted category (female Asian in this example), and a fully adapted category (male Caucasian). Each trial was followed by a 12 sec top-up adaptation period. Each target category was tested in a separate session, and in each session we measured performance both before and after adaptation. All combinations of adaptors and targets were tested for each subject. In the 2IFC RSVP task, subjects were asked to indicate which interval contained a face belonging to the target category. Masks consisted of faces belonging to the remaining three categories. Percent correct was measured as a function of presentation rate. In the VS task, 4 faces appeared simultaneously: the target and faces from the remaining three categories. Subjects reported the spatial location of the target. Response time and accuracy was recorded. Results: We found no evidence that adaptation affects saliency. We saw no effects of adaptation between fully adapted and the non-adapted targets, in either task, both within and across subjects.