DocumentCode :
3129233
Title :
Improved robust speaker identification in noise using auditory properties
Author :
Guangrui, Hu ; Xiaodong, Wei
Author_Institution :
Dept. of E.E., Shanghai Jiaotong Univ., China
fYear :
2001
fDate :
2001
Firstpage :
17
Lastpage :
19
Abstract :
The paper presents a novel approach to improve speaker identification in noise with the aid of auditory properties. The redundancy of auditory representation of speech helps to introduce the missing data technique to robust speaker identification and the auditory masking threshold is used to detect the missing parts of speech features
Keywords :
audio signal processing; noise; speaker recognition; speech intelligibility; auditory masking threshold; auditory properties; auditory representation; missing data technique; missing parts of speech features; noise; robust speaker identification; speech representation; Equations; Hidden Markov models; Humans; Loudspeakers; Masking threshold; Noise robustness; Pattern classification; Redundancy; Speech coding; Speech processing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Multimedia, Video and Speech Processing, 2001. Proceedings of 2001 International Symposium on
Conference_Location :
Hong Kong
Print_ISBN :
962-85766-2-3
Type :
conf
DOI :
10.1109/ISIMP.2001.925319
Filename :
925319
Link To Document :
بازگشت