Abstract
In tradition probability statistics model, speaker verification threshold is instability in different test situations. A novel speaker verification method based on Support Vector Data Description (SVDD) is proposed to remedy the defect of probability statistics model. To simplify the threshold value setting and improve the robustness and recognition accuracy of the verification system, traditional hard decision of SVDD is replaced by a new soft decision based on the sample acceptance rate to normalize the confidence scores to the value [0,1]. In experiment, speaker verification system based on SVDD and Gaussian Mixture Model(GMM) are compared using different length of training speech; then the system performance based on SVDD is test introducing outlier samples in training process. Experiments show that SVDD can outperform GMM, and when the target samples are not sufficient, introducing outlier samples in SVDD training process can further improve system performance.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Digital Content Technology and its Applications
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
