Face recognition by a new type of neural networks

被引：0

作者：

Ososkov, G. ^{[1
]}

Stadnik, A. ^{[1
]}

机构：

[1] Lab. of Information Technologies, Joint Institute for Nuclear Research, 141980 Dubna, Moscow region, Russia

来源：

Advances in Neural Networks and Applications | 2001年

关键词：

Algorithms - Backpropagation - Error correction - Face recognition - Image processing - Multilayer neural networks - Problem solving;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

New type of neural networks is proposed. They are suitable for direct handling of digitized images of human faces and need much less time for training than ordinary multilayer perceptrons with back propagation algorithm. Examples of efficient applications to recognizing of human faces and hand-written letters are given. The tolerance ability to various image distortions was tested.

引用

页码：304 / 308