DocumentCode
1740084
Title
Studies on a speech signal testing method based on adaptive filtering feedback technology
Author
Zhi, Chang ; Min-Tai, Wu ; Shi-Bo, Guo ; Li, Xiong Qiang
Author_Institution
Taiyuan Univ. of Technol., China
Volume
1
fYear
2000
fDate
2000
Firstpage
543
Abstract
According to the transmission mechanism of alternating speech signals, this paper puts forward a new detection method of speech signals in the real working state which is suitable for a strong-coupled acoustic wave. On the basis of the interference principle, the collected coupled acoustic signal is processed by adaptive filtering feedback technology to detect and separate the necessary speech signal. As an example, a detection test is performed to men´s and women´s speech signals which exist in an intensely coupled acoustic background in the real working state. From the comparison of the result to another result tested in an anechoic chamber, we know that this method can greatly cancel the interferences of the background acoustic wave to the tested speech signals. Thus a new experimental way which is to get speech data in an intensely coupled sound wave is realized
Keywords
acoustic signal detection; acoustic signal processing; adaptive filters; adaptive signal processing; feedback; filtering theory; noise abatement; signal detection; speech processing; adaptive filtering feedback technology; alternating speech signals; anechoic chamber; background acoustic wave; coupled acoustic background; coupled acoustic signal; interference cancellation; interference principle; speech detection test; speech signal detection method; speech signal testing method; speech transmission; strong-coupled acoustic wave; Acoustic signal detection; Acoustic testing; Acoustic waves; Adaptive filters; Adaptive signal detection; Couplings; Feedback; Interference; Signal processing; 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.2000.894549
Filename
894549
Link To Document