Discrete all-pole modeling using higher-order spectra

Discrete all-pole modeling using higher-order spectra (DAPHOS) is presented as a method for fitting all-pole envelopes to harmonic or discrete bispectra encountered in the analysis of voiced segments of speech. DAPHOS modeling combines the superior spectral envelope estimation capabilities of the DAP modeling with the Gaussian noise suppression properties of higher-order spectra. Consequently, DAPHOS outperforms DAP modeling in environments with additive Gaussian noise (white or colored) even when the noise parameters are unknown. The performance of DAPHOS modeling is compared to that of DAP modeling in both noisy and noise-free environments.<<ETX>>

[1]  Amro El-Jaroudi,et al.  Discrete all-pole modeling , 1991, IEEE Trans. Signal Process..

[2]  Chrysostomos L. Nikias,et al.  Time delay estimation in unknown Gaussian spatially correlated noise , 1988, ICASSP-88., International Conference on Acoustics, Speech, and Signal Processing.

[3]  M.R. Raghuveer,et al.  Bispectrum estimation: A digital signal processing framework , 1987, Proceedings of the IEEE.

[4]  Chrysostomos L. Nikias,et al.  The complex cepstrum of higher order cumulants and nonminimum phase system identification , 1988, IEEE Trans. Acoust. Speech Signal Process..