Improved SVDD for Speech Recognition and Simulation

Authors

Hao Rui, 1. College of Information Management, Shanxi University of Finance & Economics, Taiyuan 030006, China; ;
Xiaofeng Liu, 2. Taiyuan University of Technology, Taiyuan 030024, China;
Yanbo Niu, 2. Taiyuan University of Technology, Taiyuan 030024, China;
Xiu Lei, 1. College of Information Management, Shanxi University of Finance & Economics, Taiyuan 030006, China; ;

Abstract

Abstract: Support vector data description (SVDD) defines multi-class data by their respective hyper-spheres. The computational complexity of the quadratic programming problem is reduced significantly and it is easier to solve multi-class classification problems. Thus, SVDD has attracted more and more attention in the field of speech recognition research. For the problems of the feature vectors of speech samples overlapping and updating, the conventional SVDD for multi-class classification was improved. On the one hand, the spatial position of the samples was fully used to construct the decision function in overlapping domain of hyper-spheres; On the other hand, based on class incremental learning the dynamic change of support vectors was implemented. Simulation experimental results indicate that the proposed method reduces modeling time obviously and has better recognition performance.

Recommended Citation

Rui, Hao; Liu, Xiaofeng; Niu, Yanbo; and Lei, Xiu (2020) "Improved SVDD for Speech Recognition and Simulation," Journal of System Simulation: Vol. 29: Iss. 5, Article 11.
DOI: 10.16182/j.issn1004731x.joss.201705011
Available at: https://dc-china-simulation.researchcommons.org/journal/vol29/iss5/11

First Page

1014

Revised Date

2017-02-05

DOI Link

https://doi.org/10.16182/j.issn1004731x.joss.201705011

Last Page

1020

CLC

TP391.9

Recommended Citation

Hao Rui, Liu Xiaofeng, Niu Yanbo, Xiu Lei. Improved SVDD for Speech Recognition and Simulation[J]. Journal of System Simulation, 2017, 29(5): 1014-1020.

DOI

10.16182/j.issn1004731x.joss.201705011