基于Gabor特征和混合高斯模型的人脸表情分析

Facial expression analysis based on Gabor features and Gaussian Mixture Model

首先对人脸表情的特点进行分析,提出了利用Gabor小波特征、主分量分析(PCA)结合混合高斯模型的人脸表情分析方法,并在人脸表情数据库JAFFE进行了实验。通过对不同表情的分布规律进行实验分析,实现了对表情的定性/定量分析。实验结果表明,提出的人脸表情分析方法能够对人脸表情进行恰当的表达和描述。

This paper analyzes the nature of the facial expressions and proposes a new facial expression analysis method using Gaussian Mixture Model（GMM） based on Gabor wavelet and principal component analysis.First,the GMM parameters of each facial expression are formed.The probability of every input facial image belongs to one facial expression class can be computed based on the GMM.Then the expression analysis and recognition of facial images is achieved.The experiments on a facial expression database JAFFE have been conducted.The distribution of different expressions is analyzed,and the qualitative and quantitative description of the facial expression is accomplished.The experimental results show that the facial expression analysis method proposed in this paper can rationally represent the daily facial expressions.