Kl. Oehler et Rm. Gray, COMBINING IMAGE COMPRESSION AND CLASSIFICATION USING VECTOR QUANTIZATION, IEEE transactions on pattern analysis and machine intelligence, 17(5), 1995, pp. 461-473
Statistical clustering methods have long been used for a variety of si
gnal processing applications,including both classification and vector
quantization for signal compression. We describe a method of combining
classification and compression into a single vector quantizer by inco
rporating a Bayes risk term into the distortion measure used in the qu
antizer design algorithm. Once trained, the quantizer can operate to m
inimize the Bayes risk weighted distortion measure if there is a model
providing the required posterior probabilities, or it can operate in
a suboptimal fashion by minimizing only squared error. Comparisons are
made with other vector quantizer based classifiers, including the ind
ependent design of quantization and minimum Bayes risk classification
and Kohonen's LVQ, A variety of examples demonstrate that the proposed
method can provide classification ability close to or superior to LVQ
while simultaneously providing superior compression performance.