Application of a Novel Neural Network to Face Recognition Based on DWT

Author

Zhang, Jin ; Lou, Zhenguo ; Li, Guang ; Freeman, Walter J.

Author_Institution

Dept. of Biomed. Eng., Zhejiang Univ., Hangzhou

fYear

2006

fDate

20-22 Feb. 2006

Firstpage

1042

Lastpage

1046

Abstract

Based on studies of the biological olfactory system, a chaotic neural network model called a K-set has been constructed. This nonlinear neural network is trained to sustain a landscape of chaotic attractors for use in signal processing. Based on the idea of image partition and the discrete wavelet transform (DWT), this paper applies the chaotic neural network to face recognition. Experimental results show that the chaotic model performs image pattern classification efficiently

Keywords

chaos; discrete wavelet transforms; face recognition; image classification; learning (artificial intelligence); neural nets; pattern clustering; K-set models; biological olfactory system; chaotic neural network model; discrete wavelet transform; face recognition; image pattern classification; neural network training; nonlinear neural network; signal processing; Artificial neural networks; Biological neural networks; Biological system modeling; Brain modeling; Chaos; Discrete wavelet transforms; Face recognition; Feature extraction; Neural networks; Olfactory;

fLanguage

English

Publisher

ieee

Conference_Titel

Biomedical Robotics and Biomechatronics, 2006. BioRob 2006. The First IEEE/RAS-EMBS International Conference on

Conference_Location

Pisa

Print_ISBN

1-4244-0040-6

Type

conf

DOI

10.1109/BIOROB.2006.1639229

Filename

1639229