Title :
New Method to Analysis Noise in Speech Signal
Author_Institution :
Coll. of Electron. Eng., Guangxi Normal Univ., Guilin, China
Abstract :
The universal method to analysis non-stationary signal in reality, and the analysis methods are based on time-frequency domain or time-scale domain. An invertible transform method based on time series is proposed in this paper from another point of view. Using this method to analyze the broadband noise signal, experiments show that this method can turn disorder to order in time domain. And after transform, their characters distinguish noise signal, speech signal from each other. The principle is simple, easy to achieve.
Keywords :
noise; speech processing; time series; transforms; broadband noise signal; invertible transform method; noise analysis; nonstationary signal analysis; speech signal; time series; 1f noise; Acoustic distortion; Acoustic noise; Colored noise; Frequency; Noise level; Noise shaping; Signal analysis; Speech analysis; Speech enhancement; broadband noise; invertible transofrm; non-stationary signal;
Conference_Titel :
Circuits, Communications and Systems, 2009. PACCS '09. Pacific-Asia Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-0-7695-3614-9
DOI :
10.1109/PACCS.2009.155