一种新的嗅觉神经模型及其在人脸识别中的应用  

Novel Olfactory Neural Model and Its Application to Face Recognition

作　　者：Eduardo G.Moreira 李光[3] Yusely G.Ruiz 张锦[1] 

机构地区：[1]浙江大学生物医学工程与仪器科学学院,浙江杭州310027 [2]2.Center for Studies on [3]浙江大学工业控制技术国家重点实验室,浙江杭州310027

出　　处：《江南大学学报（自然科学版）》2008年第4期382-386,共5页Joural of Jiangnan University (Natural Science Edition)　

基　　金：国家973计划项目(2004CB720302);国家自然科学基金项目(60421002)

摘　　要：提出了一种新的嗅觉神经模型.该模型由嗅球层和嗅觉皮层构成,不同层间由前向和后向反馈构成,具有分布式延时输入,分别用二阶和一阶微分方程描述嗅球层和嗅皮层中神经元.基于ORL数据库,模型被用于人脸识别以分析其模式识别能力.实验结果表明,相比于其他算法,提出的模型具有较好的性能.A new olfactory neural model is proposed. The model has distributed delay input and is composed of two layers： olfactory bulb layer and olfactory cortex layer. There are forward and backward feedback between two layers. Neurons in olfactory bulb layer and olfactory cortex layer are described by second-order and nonlinear first-order differential equations separately. Based on ORL database, the model is used to face recognition to analyze its pattern recognition capability. The experimental results show that the proposed model has good performance compared with other algorithms.

关 键 词：嗅觉神经网络 嗅球 嗅皮层 人脸识别 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]