学习矢量量化神经网络在人脸朝向识别中的应用  

LVQ Neural Network Application in the Facial Orientation Recognition

作　　者：李萍 LI Ping(Hunyuan Normal School Attach to Datong University, Hunyuan 037400, Chin)

机构地区：[1]山西大同大学浑源师范分校,山西浑源037400

出　　处：《忻州师范学院学报》2018年第2期59-61,共3页Journal of Xinzhou Teachers University

摘　　要：人脸朝向识别的研究具有重要意义,建立了基于人工神经网络(ANN)的人脸朝向识别模型,根据学习矢量量化(LVQ)神经网络的原理,利用MATLAB神经网络工具箱,完成网络的建立、网络训练和网络测试,并通过仿真实验来实现人脸定位识别。实验结果表明,LVQ神经网络模型判断并识别任何特定人脸图像的方法是有效的。Face towards the recognition is important step in face applied research, in many research fields, face toward the identification is an unavoidable problem in pattern recognition. According to the principle of LVQ neural network, using the MATLAB neural network toolbox design creation, network training and network testing and simulation experiment, face toward the recognition. The experimental results show that the LVQ neural network model can be used to judge and identify the faces of any given face. This method is highly effi- cient and has some robustness to changes in expression and attitude.

关 键 词：人脸朝向识别 人工神经网络 学习矢量量化 特征提取 

分 类 号：TP-183[自动化与计算机技术]