Abstract
We examined the effect of facial motion in face processing by means of observer's eye movements. Face images were produced by morphing between neutral and expressive faces. A complete morph sequence lasted for 3360 ms. We created three additional sub-sequences culminating in 10%, 30% and 50% morphs, lasting for 336ms, 1008ms and 1680ms respectively. Morphing sequences and static faces were presented for identical durations. We used a 2 × 2 × 4 factorial design - presentation formats (dynamic or static), facial expressions (happy or sad) and final morph value 10%, 30%, 50% or 100%. A 2AFC task was employed. We monitored eye movements during tasks, and analyzed the data by means of FFS (Fixation-point of First Saccade) and VVF (Variance Value of Fixation-points). Our results showed that FFS was directed to centre of the face under all conditions. Moreover, VVF decreased for the moving faces in comparison to the static ones. This result suggests that facial motion may inhibit eye movement. Futhermore, VVF for happy faces was smaller than that for the sad ones. These results imply that facial motion affects eye movements after first-saccade. Maurer et al. (2002) have suggested that configural processing of faces can be divided into three type: first-order relations, holistic processing, and second-order relations. FFS could reflect acquisition of first-order relation, while VVF may reflect the analysis of second-order relations.
This work was supported by RARC(Rikkyo Amusement Research Center) and JSPS (16330144).