Title :
A Research on Recognition of Tibetan Speakers Based on MFCC and Delta Features
Author_Institution :
China Minorities Inf. Technol. Inst., Northwest Univ. for Nat. LanZhou, Lanzhou, China
Abstract :
Through experiment, the paper analyses and extracts MFCC and delta features of different Tibetan speakers and makes comparison. It shows that MFCC has certain stability between frames of the same speaker, while delta feature discriminates speakers better than MFCC does.
Keywords :
speaker recognition; Mel-frequency cepstral coefficients; Tibetan speaker recognition; delta features; Cepstral analysis; Character recognition; Data mining; Ear; Feature extraction; Linear predictive coding; Loudspeakers; Mel frequency cepstral coefficient; Speaker recognition; Speech recognition; Delta fearures; MFCC; Recognition; Tibetan;
Conference_Titel :
Computer Science-Technology and Applications, 2009. IFCSTA '09. International Forum on
Conference_Location :
Chongqing
Print_ISBN :
978-0-7695-3930-0
Electronic_ISBN :
978-1-4244-5423-5
DOI :
10.1109/IFCSTA.2009.178