Variable rate self organizing neural networks for video compression

Author

Thyagarajan, K.S. ; Erickson, Daniel

Author_Institution

Dept. of Electr. & Comput. Eng., San Diego State Univ., CA, USA

Volume

1

fYear

1994

fDate

31 Oct-2 Nov 1994

Firstpage

244

Abstract

This paper describes a design of Kohonen´s self-organizing neural networks as learning vector quantizers (LVQ) to compress video images. Both fixed rate and variable rate LVQs have been designed. For fixed rate LVQs, both full search and tree-structured codebooks are designed. Further, this paper describes the design of variable rate LVQs. Variable rate LVQs, structured as unbalanced trees, are found to provide improved performance of up to 3 dB peak SNR over comparable fixed rate LVQs

Keywords

self-organising feature maps; vector quantisation; video coding; Kohonen´s self-organizing neural networks; SNR; design; fixed rate; full search codebooks; learning vector quantizers; performance; tree-structured codebooks; unbalanced trees; variable rate self organizing neural networks; video compression; video images; Artificial neural networks; Filtering; Image coding; Neural networks; Organizing; Pattern recognition; Pixel; Self-organizing networks; Signal to noise ratio; Video compression;

fLanguage

English

Publisher

ieee

Conference_Titel

Signals, Systems and Computers, 1994. 1994 Conference Record of the Twenty-Eighth Asilomar Conference on

Conference_Location

Pacific Grove, CA

ISSN

1058-6393

Print_ISBN

0-8186-6405-3

Type

conf

DOI

10.1109/ACSSC.1994.471453

Filename

471453