Title :
Nonlinear modeling and processing of speech based on sums of AM-FM formant models
Author :
Lu, Shan ; Doerschuk, Peter C.
Author_Institution :
Sch. of Electr. & Comput. Eng., Purdue Univ., West Lafayette, IN, USA
fDate :
4/1/1996 12:00:00 AM
Abstract :
We describe a new statistical approach based on nonlinear filtering ideas for decomposing signals that are modeled as a sum of jointly amplitude- and frequency-modulated cosines, where each cosine has a slowly varying center frequency and the sum of terms is observed in additive noise. This is an alternative approach to methods based on deterministic models such as the Kaiser-Teager (see Proc. IEEE ICASSP-93, vol.III, p.149 and IEEE Trans. Acoust., Speech, Signal Processing, vol.28, no.5, pp. 599, 1980) energy operator. The Cramer-Rao bound for the resulting statistical estimation problem is computed. A practical nonlinear filter, an extended Kalman filter, is described. We demonstrate the ideas on a variety of speech problems
Keywords :
Kalman filters; amplitude modulation; filtering theory; frequency modulation; noise; nonlinear filters; parameter estimation; speech processing; statistical analysis; AM-FM formant models; Cramer-Rao bound; additive noise; amplitude-modulated cosines; extended Kalman filter; frequency-modulated cosines; nonlinear filter; nonlinear filtering; nonlinear speech modeling; nonlinear speech processing; slowly varying center frequency; speech problems; statistical estimation; sums; Additive noise; Amplitude modulation; Band pass filters; Bandwidth; Filtering; Frequency; Nonlinear filters; Phase modulation; Speech analysis; Speech processing;
Journal_Title :
Signal Processing, IEEE Transactions on
