Facial Expression Recognition Using PCA and AdaBoost Algorithm
Wang Guojiang1, and
Yang Guoliang2
1.College of Control Engineering, Chengdu University of Information Technology, Chengdu, 610225, China
2.School of Mechanical and Electrical Engineering, Jiangxi University of Science and Technology, Ganzhou, 341000, China
2.School of Mechanical and Electrical Engineering, Jiangxi University of Science and Technology, Ganzhou, 341000, China
Abstract—Facial expression recognition has an important position in affective computing. Due to the robustness of Gabor features against local distortions caused by variance of illumination, expression and pose, they have been successfully applied for face recognition. In order to effectively reduce the feature redundancy of Gabor features, in this paper, a combined classifier based on PCA and AdaBoost algorithm is proposed to recognize facial expressions. Each PCA feature vector is regarded as a projection space, and a series of weak classifiers are trained respectively. Then, the Adaboost algorithm is used to find a subset with the best classification performance from this series of weak classifiers. Finally, the PCA feature vector corresponding to this subset is used to form a new projection space, and the training samples are re-dimensionalized and a BP network classifier is trained. The experiment results show that the performance of this approach is feasible and has better result than normal method.
Index Terms—gabor wavelets transform, PCA, AdaBoost, facial expression recognition
Cite: Wang Guojiang and Yang Guoliang, "Facial Expression Recognition Using PCA and AdaBoost Algorithm," International Journal of Signal Processing Systems, Vol. 7, No. 2, pp. 73-77, June 2019. doi: 10.18178/ijsps.7.2.73-77
Index Terms—gabor wavelets transform, PCA, AdaBoost, facial expression recognition
Cite: Wang Guojiang and Yang Guoliang, "Facial Expression Recognition Using PCA and AdaBoost Algorithm," International Journal of Signal Processing Systems, Vol. 7, No. 2, pp. 73-77, June 2019. doi: 10.18178/ijsps.7.2.73-77