计算机技术与发展2018，Vol.28Issue(6)：17-20,4.DOI:10.3969/j.issn.1673-629X.2018.06.004
卷积网络的无监督特征提取对人脸识别的研究
Research on Unsupervised Feature Extraction Based on Convolutional Neural Network for Face Recognition
摘要
Abstract
At present,the learning method based on convolutional neural network needs a large number of labeled data. In practical appli-cations,it is very difficult to mark large amounts of data. In order to solve this problem,we propose an unsupervised feature extraction method based on convolutional neural network,which combines the locality preserving projection ( LPP) and the convolutional neural network. The LPP can preserve the local structure of image greatly,so it is used to learn the convolution kernel. The network structure constructed is simple and effective,and its recognition efficiency is better than the supervised convolutional neural network. The experi-ment shows that the proposed method achieves better performance than the current mainstream unsupervised feature learning methods in both real-world Yale and the classical FERET dataset.关键词
无监督特征提取/卷积神经网络/局部保持投影/人脸识别Key words
unsupervised feature extraction/convolutional neural network/locality preserving projection/face recognition分类
信息技术与安全科学引用本文复制引用
杜柏圣..卷积网络的无监督特征提取对人脸识别的研究[J].计算机技术与发展,2018,28(6):17-20,4.基金项目
国家自然科学基金(61170089) （61170089）
