Abstract
It is an open question how the dynamic property of a moving face affects the recognition of the facial expression. To solve this question, we compared moving and static morphed faces to static morphed faces through the use of four facial expressions (happy, sad, disgusted and angry). Method: Ten subjects took part in the experiment. A 4AFC task was employed to judge the facial expression. The degree of emotion of moving face images was set by morphing the original moving face on a scale of one to ten over the range from the neutral to the prototype face with a view to defining ten static face images for each emotion. The moving face images were randomly presented for all the emotions and the static face images were also presented at random. The durations of the static and moving face images were exactly the same for each morphing rate.
Results: Within the range from 30% to 50% of morphing rates of happy faces, the percentages of correct response increased for the moving faces as compared with the correct responses for the corresponding static faces. In the case of the sad face, too, the performance for the moving faces was higher than in the case of the corresponding static faces within the range from 50% to 70% of morphing rates.
Conclusion: These results suggest that the dynamic property plays an important role in the recognition of facial expression even though the morphing rate is small. This implies that the dynamic property serves as a cue for the judgment on the facial expressions at an early stage of observation. Particularly, in the case of a happy face, the dynamic property may facilitate the utilization of the low spatial frequency component of the face and hence configural information on the face.