基于密集连接卷积结构的人脸表情识别研究

Facial Expression Recognition Based on Dense Connected Convolution Structure

人脸的表情识别是图像处理当中的一个重要分支同时也是计算机视觉研究中的一个热门。采用传统卷积神经网络的人脸表情识别通过不断地加深网络层数,扩大模型参数的规模来提高识别的精度,但和传统机器学习算法对比,其精度并没有获得显著的提升。为此受到密集连接卷积神经网络(DenseNet)启发,设计了一种用于人脸表情识别的网络模型MDenseNet。使用该模型对Fer2013灰度表情识别数据集进行实验,在保证了70.45%的较高精度情况下,与传统卷积神经网络相比,模型参数的利用率更高,所需模型的参数数量大大降低。

Face expression recognition is an important branch of image processing and a hot topic in computer vision research.The traditional convolutional neural network improve the accuracy of the network by deepening the number of network layers and expanding the scale of model parameters,but compared with the traditional machine learning algorithm,its accuracy has not been significantly improved.For this reason,inspired by DenseNet,a network model M-Densenet is designed,which is specifically applied to facial expression recognition.This model is used to conduct experiments on Fer2013 gray expression recognition data set.The experimental results show that the model is used to conduct experiments on Fer2013 gray expression recognition data set.With a high accuracy rate of 70.45%,compared with the traditional convolutional neural network,the number of parameters of the model is lower and the utilization rate of model parameters is higher.