A SOLUTION TO SYSTEM ACCESS PROBLEM BASED ON CONVOLUTIONAL NEURAL NETWORK APPROACH
Abstract
The algorithm based on the principal components of the images as informative features extracting and convolutional neural network as a classifier is proposed. Experiments have shown partial invariance to the rotation, scaling and deformation, achieved through a hierarchical extraction of informative features. The resulting percentage of correct recognition rate of persons is up to 99%.
About the Author
N. A. Iskra
Белорусский государственный университет информатики и радиоэлектроники
Belarus
Belarus
References
1. Bryliuk D., Starovoitov V. // Proc. of the 2nd International Conference on Artificial Intelligence, Crimea, Ukraine, 2002. P. 428.
2. Lawrence S., Giles C.L., Tsoi A.C. et. al. Face Recognition: A Hybrid Neural Network Approach // IEEE Trans. on Neural Networks. 1997. Vol. 8, №1. P. 98.
3. Хайкин С. Нейронные сети. Полный курс. Второе издание. Москва, Санкт-Петербург, Киев, 2006.
4. LeCun Y., Boser B., Denker J. // Advances in Neural Information Processing. 1999. P. 396.
5. The Database of Faces (formerly «The ORL Database of Faces»), AT&T Laboratories, Cambridge. [Электронный ресурс]. - Режим доступа: http://www.cl.cam.ac.uk/research/. - Дата доступа: 10.05.2013.
Review
For citations:
Iskra N.A. A SOLUTION TO SYSTEM ACCESS PROBLEM BASED ON CONVOLUTIONAL NEURAL NETWORK APPROACH. Doklady BGUIR. 2013;(4):74-78. (In Russ.)
Views: 302
Email this article 