GPU based Partially Connected Neural Evolutionary network and its application on gender recognition with face images
Dept. of Comput. Sci., Xiamen Univ., Xiamen, China
8th World Congress on Intelligent Control and Automation (WCICA), 2010
@conference{chen2010gpu,
title={GPU based Partially Connected Neural Evolutionary network and its application on gender recognition with face images},
author={Chen, X.X. and Shi, M.H. and de Garis, H.},
booktitle={Intelligent Control and Automation (WCICA), 2010 8th World Congress on},
pages={1930–1934},
organization={IEEE}
}
An algorithm for evolving neural network via the genetic algorithm based on GPU parallel architecture was implemented on the CUDA, resulting in a system called CuParcone (CUDA based Partially Connected Neural Evolutionary) and was used on gender face recognition. By using the powerful ability of GPU parallel computing, CuParcone achieves a performance increase about 323 times than Parcone algorithm, which runs on a single-processor. With this new model, a gender recognition experiment was made on 530 face images (265 females and 265 males from Color FERET database), including not only frontal faces but also the faces rotated from -40 degrees ~40 degrees in the direction of horizontal, and achieved the accuracy rate of 90.84%.
April 30, 2011 by hgpu
