The application research of speech feature extraction based on the manifold learning
Traditional MFCC phonetic feature will lead a slower learning speed on account of it has high dimension and is large in data quantities.In order to solve this problem,we introduce a manifold learning,putting forward a new extraction method of MFCC-Manifold phonetic feature.We can reduce dimensions by making use of ISOMAP algorithm which bases on the classical MDS (Multidimensional scaling).Introducing geodesic distance to replace the original European distance data will make twenty-four dimensional data,which using the traditional MFCC feature extraction down to two dimensional data.Experiments prove that MFCC-Manifold feature extraction methods has achieved a satisfactory effect in data volume reduction.
manifold learning MFCC-Manifold geodesic distance feature extraction
Penghao Zhang Li Wang
College of Computer Science & Information Guizhou University Guiyang,China
国际会议
杭州
英文
796-799
2013-03-22（万方平台首次上网日期，不代表论文的发表时间）