Title :
Predicting VQ Performance Bound for LSF Coding
Author :
Chatterjee, Saikat ; Sreenivas, T.V.
Author_Institution :
Inst. of Sci., Bangalore
fDate :
6/30/1905 12:00:00 AM
Abstract :
For vector quantization (VQ) of speech line spectrum frequency (LSF) parameters, we experimentally determine a mapping function between the mean square error (MSE) measure and the perceptually motivated average spectral distortion (SD) measure. Using the mapping function, we estimate the minimum bits/vector required for transparent quantization of telephone-band and wide-band speech LSF parameters, respectively, as 22 bits/vector and 36 bits/vector, where the distribution of LSF vector is modeled as a Gaussian mixture model (GMM).
Keywords :
Gaussian processes; mean square error methods; speech coding; vector quantisation; Gaussian mixture model; mapping function; mean square error; spectral distortion; speech line spectrum frequency parameters; telephone-band speech LSF parameters; transparent quantization; vector quantization; wide-band speech LSF parameters; Bit rate; Distortion measurement; Extrapolation; Frequency measurement; Mean square error methods; Predictive models; Rate-distortion; Speech coding; Vector quantization; Wideband; Gaussian mixture model; line spectrum frequency (LSF) quantization; vector quantization;
Journal_Title :
Signal Processing Letters, IEEE
DOI :
10.1109/LSP.2007.914786
