DocumentCode
1739533
Title
Reference point alignment frequency warp method for speaker adaptation
Author
LEE, Tranzai ; Zheng, Fang ; Wu, Wenhu
Author_Institution
Dept. of Comput. Sci. & Technol., Tsinghua Univ., Beijing, China
Volume
2
fYear
2000
fDate
2000
Firstpage
756
Abstract
The variations of speakers´ vocal tract shapes result in the variations of the formant positions and sequentially in the variances of the features extracted from every frame of speech. In order to remove or reduce the variations of the formant positions, a speaker adaptation method is proposed and investigated in this paper which is based on a frequency warp function (f.w.f.). The f.w.f. warps the frequency axis so that the variations can be reduced. For a given speaker, some frequency reference points are selected to help to get this f.w.f. by finding the relationship between the positions of these reference points before and after the warping. According to the new positions of those reference points for the given speaker, the f.w.f. can then be constructed. The experimental results show that this method reduces the error rate by an average of 14.5%
Keywords
feature extraction; speech recognition; continuous speech recognition; error rate reduction; feature extraction; formant positions; frequency axis; frequency reference points; frequency warp function; reference point alignment frequency warp method; speaker adaptation; speech frame; vocal tract shapes; Computer science; Error analysis; Feature extraction; Frequency; Hidden Markov models; Intelligent systems; Laboratories; Maximum likelihood linear regression; Shape; Speech;
fLanguage
English
Publisher
ieee
Conference_Titel
Signal Processing Proceedings, 2000. WCCC-ICSP 2000. 5th International Conference on
Conference_Location
Beijing
Print_ISBN
0-7803-5747-7
Type
conf
DOI
10.1109/ICOSP.2000.891621
Filename
891621
Link To Document