| Summary: | This study used 3D distance vector measurements as the facial feature to classify six basic expressions and the distance vectors are chosen
based on Facial Action Coding System (FACS) component, facial action units (AUs).The statistical values are calculated and analyze to determine the AUs involved in facial expression and distance vectors to be taken into account to measure the intensity of each facial expression in a quantitative manner.As a result, 14 facial points are classified as significant in facial expression classification.Those facial points are in the eye, eyebrow and mouth region only. This work reveals that it is not necessary to rely on all facial feature points in estimating facial expression intensity.For Sad expression, the random mean and standard deviation of distance measurements do not indicate which AU should be taken into account to classify this expression.
|
|---|