A Study on the Pitch Search Time Reduction of Cepstrum Method 


Vol. 34,  No. 10, pp. 310-316, Oct.  2009


PDF
  Abstract

The accurate pitch extraction is very important in speech signal processing. If we measure the pitch period accurately, the accuracy of speech recognition can be higher due to the decrement of speaker dependent effect and we can change the characteristic of synthetic voice easily. Cepstrum pitch search method is a good pitch search method for transition region of speech but it has a drawback of excessive processing time for transform of processing domain. In this paper, we proposed a new method that can reduce the pitch search time of Cepstrum method. To improve the excessive processing time of conventional method for transformation of processing domain, we use the method that omits the bit-reversing process of FFT and IFFT and searches pitch peak only between 18 and 142 quefrency. As a result of applying the proposed method, we can reduce the pitch searching time by 87.94% over the conventional method.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

W. Jo, S. Choi, M. Bae, "A Study on the Pitch Search Time Reduction of Cepstrum Method," The Journal of Korean Institute of Communications and Information Sciences, vol. 34, no. 10, pp. 310-316, 2009. DOI: .

[ACM Style]

Wang-Rae Jo, Seong-Young Choi, and Myung-Jin Bae. 2009. A Study on the Pitch Search Time Reduction of Cepstrum Method. The Journal of Korean Institute of Communications and Information Sciences, 34, 10, (2009), 310-316. DOI: .

[KICS Style]

Wang-Rae Jo, Seong-Young Choi, Myung-Jin Bae, "A Study on the Pitch Search Time Reduction of Cepstrum Method," The Journal of Korean Institute of Communications and Information Sciences, vol. 34, no. 10, pp. 310-316, 10. 2009.

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880