Fig. 2From: Facial expression recognition using three-stage support vector machinesTwo-stage model: first stage (21 SVM) and Second and third stages combining the results of Stage 1Back to article page