"Quantizers for the Gamma Distribution and Other
Symmetrical Distributions", IEEE Trans. Acoustics, Speech, Signal
Processing, vol. 32, no. 4, pp. 836-841, Aug. 1984.
"Quantizers for the Gamma Distribution and Other
Symmetrical Distributions", IEEE Trans. Acoustics, Speech, Signal
Processing, vol. 32, no. 4, pp. 836-841, Aug. 1984.
This paper discusses minimum mean-square error quantization for symmetric distributions. If the distribution satisfies a log-concavity condition, the optimal quantizer is itself symmetric. For the gamma distribution often used to model speech signals, the log-concavity condition is not satisfied. It is shown that for this distribution both the uniformly spaced and the nonuniformly spaced optimal quantizers are not symmetrical for even numbers of quantizer levels. New quantization tables giving the optimal levels for quantizers for the gamma distribution are presented. A simple family of symmetric distributions is also examined. This family shows that as the distribution gets concentrated near the point of symmetry, nonsymmetric solutions become optimal.
"Computational Considerations in Adaptive
Transform Coding of Speech", Proc. Biennial Symp. Commun.
(Kingston, ON), pp. B5.9-B5.12, June 1984.
In recent years increasing emphasis is being placed on the digital encoding of speech and other analog signals. Adaptive transform coding (ATC) of speech signals offers high quality signal reproduction at low to moderate transmission rates (8-16 kb/s). In this paper, computational aspects of ATC are examined with the goal of developing an algorithm which produces high speech quality and can be implemented in practical real-time hardware.
"Vector Quantization in Adaptive Predictive Coding of
Speech", Proc. Biennial Symp. Commun. (Kingston, ON), pp.
B5.1-B5.4, June 1984.
The design and implementation of vector quantizers have recently attracted considerable attention in the speech coding field. In this paper, vector quantization is applied in an adaptive predictive coder, both to code the parameters of the linear predictor and to code the residual signal. Traditional speech coders have applied scalar quantization, i.e. coefficient by coefficient quantization, to these quantities.
M. Berouti, P. Kabal, and P. Mermelstein
"Digital Coding of Speech Signals",
Proc. IEEE Int. Symp. Circuits, Systems (Montreal, QC), pp. 6-9, May
1984.
This paper gives an overview of current work in digital speech coding. New directions in research on speech coding algorithms are discussed. At medium rates, new algorithms which use perceptually motivated fidelity criteria have significantly improved the quality of the reproduced speech. At lower rates, the application of vector quantization concepts have resulted in even lower data rates for the same speech quality. The availability of single chip signal processors has dramatically reduced the cost of implementation of practical speech coders. At the same time, considerations of issues relating to the integration of speech coders into telephone networks have resulted in new efforts at developing international standards for speech coders.
M. Berouti, H. Garten, P. Kabal, and P. Mermelstein
"Efficient Computation and Encoding of the
Multipulse Excitation for LPC", Proc. IEEE Int. Conf. Acoustics, Speech,
Signal Processing (San Diego, CA), pp. 384-387, March 1984.
This paper discusses the analysis techniques used to derive the excitation waveform for multipulse coding of speech. A computationally efficient formulation is derived for both covariance and correlation type analyses. These methods differ in the way block edges are treated. Several methods for pulse amplitude and position determination are given, ranging from a purely sequential one to one which reoptimizes pulse amplitudes at each step. It is shown that the reoptimization scheme has a nested structure that allows a reduction in the computations. An efficient method for pulse position coding is given. This method can essentially achieve the entropy limit for randomly placed pulses. Experimental results are given for typical configurations including computational requirements and speech quality assessments.
Paper titles.