On the Vector Quantization by Means of Recurrent Neural Networks - Collections | Kyushu University Library

＜departmental bulletin paper＞
On the Vector Quantization by Means of Recurrent Neural Networks

Creator	Creator Name Ma, Ni Affiliation Affiliation Name Department of Computer Science and Communication Engineering, Research Student
	Creator Name Nishi, Tetsuo 西, 哲生ニシ, テツオ Affiliation Affiliation Name Department of Computer Science and Communication Engineering, Kyushu University 九州大学大学院システム情報科学研究院
	Creator Name Wei, Gang Affiliation Affiliation Name South China University of Technology
Language	English
Publisher	九州大学大学院システム情報科学研究院
Publisher	Faculty of Information Science and Electrical Engineering, Kyushu University
Date	1997-12-22
Source Title	Research reports on information science and electrical engineering of Kyushu University
Vol	3
Issue	1
First Page	1
Last Page	6
Publication Type	Version of Record
Access Rights	open access
JaLC DOI	https://doi.org/10.15017/1498311
Related DOI	https://portal.isee.kyushu-u.ac.jp/
Related URI	https://portal.isee.kyushu-u.ac.jp/
Relation	https://portal.isee.kyushu-u.ac.jp/
Abstract	A vector quantization scheme with a two-stage neural network coding(NNVQ) is developed, where an encoded vector is approximated by the output of the networks driven by a vector from an excitation code...book. The networks and the codebook are optimized to overcome some difficulties in conventional algorithms. The experimental results show that the NNVQ which employs the recurrent neural networks and the optimized learning algorithm performs the best among the reference versions of the VQ algorithms.show more

Hide fulltext details.

File	FileType	Size	Views	Description
p001	pdf	541 KB	148

Details

PISSN	1342-3819
EISSN	2188-0891
NCID	AN10569524
Record ID	1498311
Peer-Reviewed	Refereed
Subject Terms	Vector quantization
	Recurrent neural networks
	Learning
	Optimization
Created Date	2015.04.09
Modified Date	2020.10.15