Expression Recognition Algorithm Based on SURF and Improved AlexNet
In view of the imperfect feature information extraction and poor classification effect of traditional convolutional neu-ral network in facial expression recognition,a dual channel expression recognition algorithm combining SURF algorithm and im-proved AlexNet network is proposed in this paper.This method is based on the traditional AlexNet network.Firstly,the improved AlexNet convolutional neural network is used to extract and describe the feature points of the face image,and the extracted features are used as the main feature points.Then,the SURF algorithm is used to extract the facial features as the secondary feature points,supplement the main feature points and analyze the parameter changes.Secondly,the feature information extracted by two channels is fused to realize complementary advantages.Finally,the fused feature information is used as the input of SVM classifier to classify and output the expression classification results.This method is tested on Fer2013 and CK+data sets.Compared with the previous al-gorithms,its recognition rate is improved by 6.14%and 6.51%respectively.
万方数据
维普
