Application of BP Neural Network in Face Recognition

DOI

10.2991/mce-14.2014.120How to use a DOI?

Keywords

BP Network; Neural network; Image compression; Pattern recognition; Face recognition

Abstract

In this thesis, a BP neural network model is used to recognize the face direction, which established by simulating the human brain. The recognition model designed includes image compression, image sampling, input vector standardization, BP neural network and competition selection. Experimental results based On ORL face database. Then inputs them to train the neural network that consists of three layers, and the input, hidden and output layers have 42,70 and 40 neurons respectively. When training adjusts the parameters to ensure its performance and convergence speed. Lastly, a lot of the feature vectors are input to verify the model. By comparing the output and fact, it is found that the model is effective and reliable with an error rate of 6.5% merely. This demonstrates that there cognition model is simple and has a high recognition rate. Training speed and efficiency can be increased, and model classification performance can be improved obviously when these strategies are used cooperatively. Therefore, there cognition model and training strategies presented in this paper are practical.

Copyright

© 2014, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Wenyang Yu
AU  - XianWei Wu
AU  - YuBin Yang
PY  - 2014/03
DA  - 2014/03
TI  - Application of BP Neural Network in Face Recognition
BT  - Proceedings of the 2014 International Conference on Mechatronics, Control and Electronic Engineering
PB  - Atlantis Press
SP  - 537
EP  - 540
SN  - 1951-6851
UR  - https://doi.org/10.2991/mce-14.2014.120
DO  - 10.2991/mce-14.2014.120
ID  - Yu2014/03
ER  -