Research on Optimization of Facial Expression Recognition Algorithm Based on Convolutional Neural Network and Support Vector Machine

Facial expression recognition (FER), as an important branch of computer vision, has made significant progress in recent years thanks to the development of deep learning technology. This article proposes a support vector machine (SVM) facial expression recognition algorithm based on convolutional neural network (CNN) optimization, aiming to improve recognition accuracy and robustness. This method utilizes the advantages of traditional machine learning such as SVM and the automatic feature extraction characteristics of deep learning, proposes a new feature extraction and classification model, and further improves the model performance through data augmentation, hyperparameter optimization, and other means. Experimental verification shows that the algorithm exhibits good recognition accuracy and robustness on multiple publicly available datasets. This study provides new ideas for improving the performance of facial expression recognition systems, which is of great significance for promoting progress in this field.