By Topic

A high speed modified Hopfield neural network and a design of character recognition system

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Chen, L.-C. ; Inst. of Electron. Eng., Chung-Yung Christian Univ., Chung Li ; Chen, Y.-S.

A digital methodology called single error correction-single error detection (SECSED) is used to improve the learning and recall rules of the conventional Hopfield neural network. The XOR operator is used to implement the learning phase and the recall phase of the modified network. The bulk data processing of the modified Hopfield network is much faster than that of the conventional Hopfield network, because the improved method uses only XOR operations to avoid the carry-propagation of the multiplication and addition which is necessary in the learning and recall phases in the conventional Hopfield network. Because only the XOR operation is involved, it is very suitable for VLSI implementation of the neural network. Simulated results indicate that the modified Hopfield neural network can recognize each 24% noise-interfered pattern within 1 to 6 iterations with a recognition rate above 90%, while the conventional Hopfield neural network recognizes each small noise-interfered pattern with 6 to 12 iterations and a recognition rate below 5%

Published in:

Security Technology, 1991. Proceedings. 25th Annual 1991 IEEE International Carnahan Conference on

Date of Conference:

1-3 Oct 1991