Vector quantization of speech LSP parameters using trellis codes and l1-norm constraints

Authors:
Jianping Pan;Thomas R. Fischer
Affiliations:
School of Electrical Engineering and Computer Science, Washington State University, Pullman, WA;School of Electrical Engineering and Computer Science, Washington State University, Pullman, WA
Venue:
ICASSP'93 Proceedings of the 1993 IEEE international conference on Acoustics, speech, and signal processing: speech processing - Volume II
Year:
1993

Citing 3
Cited 0

Vector quantization and signal compression

Vector quantization and signal compression
Efficient vector quantization of LPC parameters at 24 bits/frame

ICASSP '91 Proceedings of the Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference
Uniform and piecewise uniform lattice vector quantization for memoryless Gaussian and Laplacian sources

IEEE Transactions on Information Theory

Quantified Score

Hi-index	0.00

Visualization

Abstract

Trellis-coded vector quantization (TCVQ) is used to encode the LSP parameters. Three intraframe encoding schemes are considered: direct encoding the LSP vectors, encoding the differences of LSP parameters, and using non-linear prediction in encoding the LSP vectors. The last encoding scheme is best, and can achieve 1 dB spectral distortion using about 27 bits for each speech frame. An alternative LSP coder is described, based on an l1-norm constrained vector quantizer. This two-stage VQ encoding scheme uses 25 bits/frame to achieve about 1 dB spectral distortion.