A novel entropy-constrained competitive learning algorithm for vector quantization

In this paper, we present a novel competitive learning algorithm for the de sign of a variable-rate vector quantizer (VQ). The algorithm, termed entrop y-constrained competitive learning (ECCL) algorithm, can achieve a near-opt imal performance subject to the average rate constraint. Simulation results show that, under the same average rate, the ECCL algorithm enjoys a better performance than that of the existing competitive learning algorithms. Mor eover, the ECCL algorithm outperforms the entropy-constrained vector quanti zer (ECVQ) (Chou et al., IEEE Trans. Acoust. Speech Signal Process. 37 (198 9) 31-42) design algorithm under the same rate constraint and initial codew ords. The ECCL algorithm is also more insensitive to the selection of initi al codewords as compared with the ECVQ algorithm. Therefore, the ECCL algor ithm can be an effective alternative to the existing variable-rate VQ desig n algorithms for the applications of signal compression. (C) 1999 Elsevier Science B.V. All rights reserved.