Table of Contents

NAME

ecvqtrain - train codebooks for entropy-constrained vector quantization (ECVQ)

SYNOPSIS

ecvqtrain [-i initialcodebookfile] [-s initialcodebooksize] [-I iterations] [-t threshold] lambda trainingfile codebookfile

DESCRIPTION

ecvqtrain trains codebooks for use in entropy-constrained vector quantization (ECVQ) (see ecvqencode(1) ). trainingfile is an input data file in DAT format which contains training vectors. codebookfile is the output file in CBK format which contains the codewords which form the codebook. Initial codewords are chosen at random, unless initialcodebookfile is specified.

lambda (a floating point value) is the Lagrangian-multiplier "rate-distortion" parameter that controls the trade-off between rate and distortion in a cost function used in the iterative descent algorithm. This cost function takes the form of J(i) = D(i) + lambda * R(i) where D(i) is the average distortion and R(i) is the average rate (estimated from codeword probabilities) for iteration i. See the paper by Chou, Lookabaugh, and Gray cited below for more details on the training algorithm.

Because the ECVQ training usually reduces the codebook size during training, the output codebook, codebookfile, may contain fewer codewords than the the initial codebook (as specified by initialcodebookfile or by the -s option).

OPTIONS

-i initialcodebookfile
String. Initial codebook. Note: one (and only one) of -i and -s must be given.
-s codebooksize
Integer. Number of codewords to create for random initial codebook. Note: one (and only one) of -i and -s must be given.
-I iterations
Integer. Traininging stops after iterations iterations of the ECVQ training algorithm through the training data. Note: one (and only one) of -t and -I must be given.
-t threshold
Float. Stop the training process when the Lagrangian cost function, J, changes by less than threshold from one iteration to the next; that is, stop if (J(i-1) - J(i))/J(i) < threshold, where J(i) is the Lagrangian cost for the ith iteration and is averaged over the entire training set. Note: one (and only one) of -t and -I must be given.

SEE ALSO

ecvqencode(1) , QccPackVQ(3) , QccPack(3)

P. A. Chou, T. Lookabaugh, and R. M. Gray, "Entropy-constrained Vector Quantization," IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. 37, pp. 31-42, January 1989.

AUTHOR

Copyright (C) 1997-2021 James E. Fowler


Table of Contents



Get QccPack at SourceForge.net. Fast, secure and Free Open Source software downloads