改进的F-score算法在语音情感识别中的应用  

Application of improvement of F-score algorithm in speech emotion recognition

作　　者：叶吉祥[1] 王聪慧[1] 

机构地区：[1]长沙理工大学计算机与通信工程学院,长沙410114

出　　处：《计算机工程与应用》2013年第16期137-141,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61170199);湖南省教育厅重点资助项目(No.11A004)

摘　　要：针对F-score特征选择算法不能揭示特征间互信息而不能有效降维这一问题,应用去相关的方法对F-score进行改进,利用德语情感语音库EMO-DB,在提取语音情感特征的基础上,根据支持向量机(SVM)的分类精度选择出分类效果最佳的特征子集。与F-score特征选择算法对比,改进后的算法实现了候选特征集较大幅度的降维,选择出了有效的特征子集,同时得到了较理想的语音情感识别效果。For the F-score feature selection algorithm can not reveal the mutual information among features, the method of re- moving the redundancy is applied to improve the F-score algorithm. Using the German emotional speech database EMO-DB, based on the extraction of speech emotion features, the paper uses the classification accuracy of SVM to choose the best feature subset. Compared with the F-score method, the improved feature selection algorithm can achieve dimension reduction substan- tially, select an effective feature subset, and obtain an ideal speech emotion recognition accuracy.

关 键 词：特征选择 F-score 互信息 支持向量机 语音情感识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]