The Journal of the Acoustical Society of Korea (한국음향학회지)

The Acoustical Society of Korea (한국음향학회)
권호 표지

Speaker Verification Using SVM Kernel with GMM-Supervector Based on the Mahalanobis Distance

Mahalanobis 거리측정 방법 기반의 GMM-Supervector SVM 커널을 이용한 화자인증 방법

  • 김형국 (광운대학교 전자융합공학과) ;
  • 신동 (광운대학교 전자융합공학과)
  • Received : 2010.01.19
  • Accepted : 2010.02.23
  • Published : 2010.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we propose speaker verification method using Support Vector Machine (SVM) kernel with Gaussian Mixture Model (GMM)-supervector based on the Mahalanobis distance. The proposed GMM-supervector SVM kernel method is combined GMM with SVM. The GMM-supervectors are generated by GMM parameters of speaker and other speaker utterances. A speaker verification threshold of GMM-supervectors is decided by SVM kernel based on Mahalanobis distance to improve speaker verification accuracy. The experimental results for text-independent speaker verification using 20 speakers demonstrates the performance of the proposed method compared to GMM, SVM, GMM-supervector SVM kernel based on Kullback-Leibler (KL) divergence, and GMM-supervector SVM kernel based on Bhattacharyya distance.

본 논문에서는 Gaussian Mixture Model (GMM)-supervector의 Mahalanobis 거리측정 방법 기반의 Support Vector Machine (SVM) 커널을 이용한 새로운 화자인증 방법을 제안한다. 제안된 GMM-supervector SVM 커널방식은 GMM 방식과 SVM 방식을 결합한 방식으로서, GMM 파라미터에 의해 형성된 화자 및 비 화자 GMM-supervectors의 화자인증 임계값을 Mahalanobis 거리측정 방법기반의 SVM 커널에 적용함으로써 화자인증 정확도를 높인다. 제안한 방식의 성능 측정을 위해 20명의 화자를 대상으로 문장독립형 화자인증 실험을 수행하여 기존에 사용되고 있는 GMM, SVM, Kullback-Leibler (KL) divergence 거리측정 방법 기반의 GMM-supervector SVM 커널, Bhattacharyya 거리측정 방법기반의 GMM-supervector SVM 커널 방식을 통한 화자인증 결과들과 비교하였다.

Keywords

References

  1. D. A. Reynolds, T. F. Quatieri, and R. B. Dunn, "Speaker verification using adapted Gaussian mixture models," Digit. Signal Process., vol. 10, no. 1-3, pp. 19-41, 2000. https://doi.org/10.1006/dspr.1999.0361
  2. W. M. Campbell, J. P. Campbell, and D. A. Reynolds, "Support vector machines for speaker and language recognition," Comput. Speech Lang., vol. 20, no. 2-3, pp. 308- 311, 2006.
  3. V. Wan, "Speaker verification using support vector machines," Ph.D. Dissertation, Univ. Sheffield, Sheffield, U.K., 2003.
  4. J. Louradour, K. Daoudi, and F. Bach, "Feature space Mahalanobis sequence kernels: application to SVM speaker verification," IEEE Trans. on Audio, Speech, and Language process., vol. 15, no. 8, pp. 2465-2475, 2007. https://doi.org/10.1109/TASL.2007.905147
  5. W. M. Campbell, D. E. Sturim, and D. A. Reynolds, "Support vector machines using GMM supervectors for speaker verification," IEEE Signal Process. Lett., vol. 13, no. 5, pp. 308-311, 2006. https://doi.org/10.1109/LSP.2006.870086
  6. C. H. You, K. A. Lee, and H. Li, "An SVM kernel with GMM-supervector based on the Bhattacharyya distance for speaker recognition," IEEE Signal Process. Lett., vol. 16, no. 1, pp. 49-52, 2009. https://doi.org/10.1109/LSP.2008.2006711