Face Expression Recognition Combining Convolutional Neural Network and ASM Algorithm

In this paper, we propose a Face Expression Recognition method Combining Convolutional Neural Network and ASM (Active Shape Model) Algorithm, which is referred to as the ASM-CNN method. In this method, the ASM algorithm is first utilized to mark the feature points of the key parts of the face in the image, and the face part is cut out from the original image and fed into a two-channel convolutional neural network to recognize the expression. We utilize this algorithm to separately identify the JAFFE, CK+, and Fer2013 datasets, the recognition rate of the algorithm on the test set of the three datasets is 92.13%, 90.96% and 80.08% respectively. And the time taken by the algorithm to reduce the loss function to 0.05 on the training set of these three datasets is 45.16 seconds, 147.33 seconds and 430.79 seconds, respectively. Compared with VGG-16 model and Reset-net model, the recognition rate of face expression recognition is improved by 9.12% to 17.17%, which proves the effectiveness of the algorithm.