摘 要:人脸图像能够表现大量生物学上的复杂信息,从人脸图像中对人物的年龄进行估计有助于机器视觉在安防、预测等方面的应用。本文提出了一种新的深度神经网络,利用卷积神经网络对人脸图像进行特征提取,结合多层自编码器实现对不同年龄层的分类。同时对提取的人脸特征进行统计,分析随人物衰老变化较大的神经元。在FG-NET 数据集上获得了较高准确率。
关键词:深度神经网络;人脸图像;年龄估计
中图分类号:TP391 文献标识码:A 文章编号:2096-4706(2019)18-0040-03
Age Estimation of Face Images Based on Deep Learning
LI Jue1,LU He2
(1.Military Representative Bureau of Naval Equipment Department in Beijing,Beijing 100076,China;2.Capital Aerospace Machinery Co.,Ltd.,Beijing 100076,China)
Abstract:Face images can represent a large amount of complex biological information. Estimating the age of human from face images is helpful for the application of machine vision in security,prediction and so on. In this paper,a new depth neural network is proposed,which uses convolution neural network to extract features from face images and combines with multi-layer self-encoder to classify different age levels. At the same time,the extracted facial features are counted and the neurons which change greatly with the aging of the characters are analyzed. High accuracy is obtained on FG-NET dataset.
Keywords:deep neural network;facial images;age estimation
参考文献:
[1] Jain Ak,Ross A,Prabhakar S. An Introduction to Biometric Recognition [J].IEEE Transactions on Circuits and Systems for Video Technology,2004,14(1):4-20.
[2] 王晓刚,汤晓鸥. 从统一子空间分析到联合深度学习:人脸识别的十年历程 [J]. 中国计算机学会通讯,2015(4):8-15.
[3] Ramanathan N,Chellappa R. Face Verification across Age Progression [C]// Computer Vision and Pattern Recognition,2005.CVPR 2005. IEEE Computer Society Conference on. S.l.:s.n.,2005:462-469.
[4] 何清,李宁,罗文娟,等. 大数据下的机器学习算法综述 [J]. 模式识别与人工智能,2014,27(4):327-336.
[5] 刘建伟,刘媛,罗雄麟. 深度学习研究进展 [J]. 计算机应用研究,2014,31(7):1921-1930+1942.
[6] Bengio Y,Lamblin P,Popovici D,et al. Greedy layer-wise training of deep networks [J]. Advances in Neural Information Processing Systems,2007,19:153-160.
[7] Fukushima K. Neocognitron:A Self_organizing Neural Network Model for a Mechanism of Pattern Recognition Unaffected by Shift in Position [J].Biological Cybernetics,1980,36(4):193-202.
[8] Hinton G E,Osindero S,Teh Y W. A fast learning algorithm for deep belief nets [J].Neural Computation,2006,18(7):1527-1554.
[9] Lanitis A. Evaluating the performance of face-aging algorithms [C]// IEEE International Conference on Automatic Face & Gesture Recognition,2009.
作者简介:
李珏(1990-),女,汉族,山东青岛人,助理工程师,硕士研究生,研究方向:武器装备信息化、机器学习;
卢鹤(1991-),男,汉族,北京人,助理工程师,硕士研究生,研究方向:软件工程。