摘要
神经网络由于具有抗干扰性和容错性,常被用在人脸识别的分类器中。但是,神经网络的训练过程花费的时间很长,训练的次数多,而且不易收敛。这里提出一种合作式的遗传神经网络作为新的分类器。使分类器的训练次数明显的减少,收敛性得到了提高。实验仿真得到了很好的效果。
The neural network (NN) is selected as a classifier to recognize face, because of its anti-jamming ability and robust ability, whereas the NN may spend much training time and training epochs,and may also get nonanticipant convergence. A cooperative genetic algorithm neural network (GA-NN) is represented as a new classifier in this paper. It reduces the training time and training epochs, and the convergence is improved. The simulation results show that is has good effect.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2006年第z3期2230-2231,共2页
Chinese Journal of Scientific Instrument
关键词
人脸识别
遗传算法
BP网络
face recognition the genetic algorithm the neural network
