Improved Facial Expression Recognition Algorithm Based on Convolutional Neural Network

Facial expression recognition is one of the important research directions in the field of computer vision. By analyzing facial expressions in images or videos to infer a person's emotional state or expression, convolutional neural networks can extract features from facial expression images in order to determine the type of expression. Technology based on facial expression recognition can be used in a variety of occasions. This paper is based on the Res-18 network. This article is based on the Res-18 network for research, using PSConv to replace the convolutional layers in the benchmark network, and conducting facial expression recognition experiments on the RAF-DB facial expression dataset. The experimental results show that the recognition rate in this dataset reached 85.14%, which is 3% higher than the benchmark network. Therefore, it can be concluded that the network with replaced convolution has better robustness.