A two stage vector quantization approach via self-organizing map
Contribuinte(s) |
Baozong, Yuan Xiaofang, Tang |
---|---|
Data(s) |
01/01/2002
|
Resumo |
In this paper, a two-stage algorithm for vector quantization is proposed based on a self-organizing map (SOM) neural network. First, a conventional self-organizing map is modified to deal with dead codebooks in the learning process and is then used to obtain the codebook distribution structure for a given set of input data. Next, subblocks are classified based on the previous structure distribution with a prior criteria. Then, the conventional LBG algorithm is applied to these sub-blocks for data classification with initial values obtained via the SOM. Finally, extensive simulations illustrate that the proposed two-stage algorithm is very effective.<br /> |
Identificador | |
Idioma(s) |
eng |
Publicador |
IEEE |
Relação |
http://dro.deakin.edu.au/eserv/DU:30044861/venkatesh-atwostage-2002.pdf http://dx.doi.org/10.1109/ICOSP.2002.1181205 |
Direitos |
2002, IEEE |
Palavras-Chave | #code standards #computer networks #convergence #data compression #distortion measurement #euclidean distance #neural networks #nonlinear distortion #organizing #vector quantization |
Tipo |
Conference Paper |