High-Rate Optimized Recursive Vector Quantization Structures Using Hidden Markov Models

This paper examines the design of recursive vector quantization systems built around Gaussian mixture vector quantizers. The problem of designing such systems for minimum high-rate distortion, under input-weighted squared error, is discussed. It is shown that, in high dimensions, the design problem...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on audio, speech, and language processing Vol. 15; no. 3; pp. 756 - 769
Main Authors Duni, E.R., Rao, B.D.
Format Journal Article
LanguageEnglish
Published Piscataway, NJ IEEE 01.03.2007
Institute of Electrical and Electronics Engineers
Subjects
Applied sciences
Coding
Coding, codes
Design engineering
Detection, estimation, filtering, equalization, prediction
Distortion
Distortion measurement
Exact sciences and technology
Frequency measurement
Hidden Markov model (HMM)
Hidden Markov models
high-rate quantization
Image coding
Information, signal and communications theory
log spectral distortion (LSD)
Mathematical models
Maximum likelihood estimation
Nonlinear distortion
Parameter estimation
Quantization
Recursive
Sampling, quantization
Signal and communications theory
Signal processing
Signal, noise
Speech
Speech processing
Telecommunications and information theory
Vector quantization
Wideband
wideband speech
Parameter estimation
Spectral method
log spectral distortion (LSD)
Probabilistic approach
Vector quantization
Data compression
Rate distortion theory
Random coding
wideband speech
Line spectrum
Hidden Markov model (HMM)
Speech coding
Gaussian process
Wide band
Signal quantization
Hidden Markov models
high-rate quantization
Recursive method
Maximum likelihood
Speech processing
Signal distortion
Online AccessGet full text

Cover

More Information
Summary:This paper examines the design of recursive vector quantization systems built around Gaussian mixture vector quantizers. The problem of designing such systems for minimum high-rate distortion, under input-weighted squared error, is discussed. It is shown that, in high dimensions, the design problem becomes equivalent to a weighted maximum likelihood problem. A variety of recursive coding schemes, based on hidden Markov models are presented. The proposed systems are applied to the problem of wideband speech line spectral frequency (LSF) quantization under the log spectral distortion (LSD) measure. By combining recursive quantization and random coding techniques, the systems are able to attain transparent quality at rates as low as 36 bits per frame
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:1558-7916
1558-7924
DOI:10.1109/TASL.2006.885903