Facial Expression Recognition from 2.5D Partial Face Data by Using Face Plane

Abstract

Due to a problem of current research occurring when recognizing facial expressions from a 2.5D partial face data set taken from any viewpoint ranging from -45 to +45, we propose a novel algorithm for recognizing facial expressions from a 2.5D partial face data set. A 2.5D partial data set is captured from any viewpoint between -45 and +45. For facial expression recognition, a 3D virtual expression face is first reconstructed from a 2.5D partial face data set. A facial expression is then represented in terms of the change of crossing points on a face plane. Next, two schemes are used to analyse the crossing points for recognition. In the ¯rst scheme, the distribution of crossing points was used. In the second scheme, the displacement vectors of crossing points by facial expression change were used. The classi¯cation is carried out by using a support vector machine (SVM) for the both recognition schemes. The experiments were done for four facial expressions (neutral, anger, surprise and smiling) of 22 persons. The results showed the feasibility of the proposed method.