基于面部和动作表情的双模态情绪强度估计  

Double-Mode Estimation of Emotion Intensity Based on Facial and Action's Expression

作　　者：李春玲[1] 邹北骥[1,2] 王磊[1] 

机构地区：[1]中南大学信息科学与工程学院图形图像处理实验室,长沙410083 [2]怀化学院计算机与科学技术学院,怀化418008

出　　处：《系统仿真学报》2009年第16期5047-5052,共6页Journal of System Simulation

基　　金：国家自然科学基金(60673093)

摘　　要：针对急躁情绪的强度估计问题,提出了解决方案,其思路是:对急躁情绪进行数学建模,构建双模态特征模型,分别采用面部表情和动作表情(头部运动)双模态信息来表征急躁情绪;然后用双层高斯混合隐马尔科夫模型(DMHMM)对情绪强度进行估计,得到情绪强度的三个等级。实验结果表明:该算法实现简单,并且能快速的对情绪强度进行估计,与只采用面部表情信息的方法相比正确率提高了24.4%,达到84.8%。A solution for the estimation of impatience intensity was proposed as follows： a mathematical modeling for impatience was advanced to build a double-mode feature model, expressing impatience with both facial and action＇s expression, and then, emotion intensity was estimated and it was divided into three grades by Double layers of Gaussian Mixture Hidden Markov Model （DMHMM）. The experiment results indicate that it is a simple and effective algorithm of intensity estimation and the accuracy is heightened by 24.4% to 84.8% compared with the method based on facial expression.

关 键 词：面部表情 动作表情 双模态 情绪强度 DMHMM 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]