基于卷积神经网络的人脸表情识别  

Facial expression recognition based on CNN

作　　者：徐鹏[1] 薄华[1] 

机构地区：[1]上海海事大学信息工程学院,上海201306

出　　处：《微型机与应用》2015年第12期45-47,共3页Microcomputer & Its Applications

摘　　要：传统的神经网络表情识别系统由特征提取和神经网络分类器组成,利用人的经验来获取模式特征,很容易丢失表征表情特征的细节信息。提出一种基于卷积神经网络的识别方法,避免了对图像进行复杂的特征提取,直接把图像数据作为输入。通过在Cohn-Kanade表情库上的实验结果表明,该方法能够取得很好的表情分类效果。Traditional neural network face recognition system consists of feature extraction and neural network classifiers, using human experience to get the pattern features. It is easy to lose face features characterizing details. This paper proposes a method based on the convolutional neural network without the complex feature extraction of the image. Results of the experiment on Cohn- Kanada show that the method described in this paper has a promising performance.

关 键 词：卷积神经网络 人脸表情识别 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]