Abstract
Previous research has shown that adapting to a face horizontally rotated about a vertical axis produces a perceptual shift, where the test face appears rotated slightly away from the direction of the adapting face (Fang & He, Neuron, 2005). We have recently confirmed this finding in our lab using synthetic face stimuli. In the current study, we sought to explore how the geometric elements of our stimuli independently contribute to this effect. In a two alternative forced choice task, subjects were presented with an adapting face oriented 20 degrees to the left or right for four seconds, followed by a briefly presented test face, which was randomly chosen in each trial from a set of seven faces spanning +/− 6° around a frontal view. Subjects were instructed to choose whether each test face appeared left or right of center. By assessing the orientation of the test face at which subjects were equally likely to choose left or right (point of subjective equality), we were able to assess the strength of adaptation. We tested subjects in three conditions: Adapting to full faces (Intact), head outlines only (Outline), and features only (Features). In all conditions, the test faces were full faces. We found that Intact adapted more strongly than Features (pOutline adapted stronger than Features (pIntact vs. Outline (p[[lt]]0.123). These results suggest a non-linear combination of outline and features, with a privileged role for the head outline in encoding the direction of gaze.
This work was supported by a CIHR grant # 172103 to H.R.W.