3d Smiling Facial Expression Recognition Based On Svm

Shuming Liu,Xiaopeng Chen,Di Fan,Xu Chen,Fei Meng,Qiang Huang
DOI: https://doi.org/10.1109/ICMA.2016.7558813
2016-01-01
Abstract:Using Kinect acquired RGB-D image to obtain a face feature parameters and three-dimensional coordinates of the characteristic parameters, and to select the characteristic parameter Facial by Candide-3 model, and feature extraction and normalization. Smile face expression data collection through Kinect, SVM collected to smiley face data classify and output the result of recognition, and the results compared with two-dimensional image of smiling face expression recognition results. Experimental results show that three-dimensional image of smiling face expression recognition accuracy than the two-dimensional image of smiling face. This research has important significance for the research and application of facial expression recognition technology.
What problem does this paper attempt to address?