Title :
Parallel architectures for vector quantization
Author :
Ancona, Fabio ; Rovetta, Stefimo ; Zunino, Rodolfo
Author_Institution :
Dept. of Biophys. & Electron. Eng., Genoa Univ., Italy
Abstract :
The paper describes a parallel implementation of neural networks based on vector quantization. A toroidal-mesh topology has been used to assess the overall approach. A theoretical analysis of the modular system´s efficiency is presented. The final application goal is a lossy compression of high-dimensional data for low bit-rate communications. Experimental results on a significant testbed shows a remarkable increase of the system´s performances. In addition, the fit between predicted and measured efficiency values confirms the validity of the overall theoretical model
Keywords :
neural net architecture; parallel architectures; vector quantisation; lossy compression; low bit-rate communications; neural networks; parallel architectures; toroidal-mesh topology; vector quantization; Computer architecture; Costs; Image coding; Image processing; Iterative algorithms; Parallel architectures; Prototypes; Signal processing algorithms; Vector quantization; Very large scale integration;
Conference_Titel :
Neural Networks,1997., International Conference on
Conference_Location :
Houston, TX
Print_ISBN :
0-7803-4122-8
DOI :
10.1109/ICNN.1997.616144
