Research Article

Facial Expression Recognition Using Stationary Wavelet Transform Features

Table 5

Comparison of the proposed facial expression recognition system with state-of-the-art methods for JAFEE dataset.

Method, yearFeaturesOverall CRR (emotions)

ProposedSWT + DCT98.83%  
[28], 2016Curvelet transform95.17%  
[27], 2011Patch based Gabor92.93%  
[26], 2005Gabor + FSLP91.00%  
[23], 2008DCT79.30%