Abstract
The aim of the current study was to address the long-standing question of whether three-
dimensional information plays a role in face recognition, as well as to test the hypothesis that the
advantage of the intermediate view derives from better structural information about the face
provided by this view. We examined these issues by comparing recognition performance across a
change in viewpoint with or without stereo information. Subjects' eye-movements were also
recorded during both 2D and 3D sessions. Our data analysis supported conclusions from the
previous studies that a 22.5° intermediate view leads to better face recognition performance
compared to the full frontal view. The results revealed that participants were more accurate in the
3D condition, while ‘region of interest’ analysis of gaze data showed that rich volumetric
properties provided by certain facial features were attended more in the 3D condition compared
to the 2D condition. Taken together, these two findings support the conclusion that face
recognition across viewpoint transformation is facilitated with the addition of stereoscopic depth
cues. Contrary to our predictions, we did not find a statistically significant interaction effect
between depth information and angle of view of the face. However, results of a follow-up
statistical analysis suggest the effect could become significant with the addition of more data.