基于深度卷积神经网络的人眼检测Eye detection based on deep convolutional neural networks
刘俊超,陈志军,樊小朝,闫学勤,王宏伟
摘要(Abstract)：
基于人眼检测的驾驶员疲劳检测研究中,受戴眼镜、光照强度和人脸姿态变化、闭眼以及面部遮挡等复杂背景影响,现有方法难以准确检测人眼位置的问题越来越严重,故提出一种基于深度卷积神经网络的人眼检测方法。对其进行网络优化及损失优化,把人眼检测作为一种回归问题来求解,实现了整个过程端到端,即从输入原始图片到最后的人眼类别和位置的输出。该方法在ORL人脸数据库中全部图像的人眼检测准确率为98.39%,在AR人脸数据库中没有戴墨镜的人眼检测准确率为95.15%,实验结果验证了所提方法的有效性、高准确率和较强的泛化能力。
关键词(KeyWords)： 人眼检测;深度学习;卷积神经网络;网络优化;损失优化;泛化能力
基金项目(Foundation): 国家自然科学基金资助项目(51666017);; 新疆维吾尔自治区自然科学基金资助项目(2015211C272;2016D01C062)~~
作者(Author): 刘俊超,陈志军,樊小朝,闫学勤,王宏伟
DOI: 10.16652/j.issn.1004-373x.2018.18.016
参考文献(References)：
- [1]国家统计局.中国统计年鉴[M].北京:中国统计出版社,2013.National Bureau of Statistics.China statistical yearbook[M].Beijing:China Statistics Press,2013.
- [2]PENG Yan,ZHOU Tian,WANG Shaopeng,et al.Design and implementation of a real-time eye tracking system[J].The journal of China universities of posts and telecommunications,2013,20(S1):1-5.
- [3]YOUNG D,TUNLEY H,SAMUELS R.Specialised hough transform and active contour methods for real-time eye tracking[J/OL].[1995-07-01].http://users.sussex.ac.uk/~davidy/eyetrack_csrp.pdf.
- [4]ZHOU Z H,GENG X.Projection functions for eye detection[J].Pattern recognition,2004,37(5):1049-1056.
- [5]SONG Fengyi,TAN Xiaoyang,LIU Xue,et al.Eyes closeness detection from still images with multi-scale histograms of principal oriented gradients[J].Pattern recognition,2014,47(9):2825-2838.
- [6]CHEN X,XIANG S,LIU C L,et al.Vehicle detection in satellite images by parallel deep convolutional neural networks[C]//Proceedings of 2nd IAPR Asian Conference on Pattern Recognition.Naha:IEEE,2014:181-185.
- [7]HARIHARAN B,ARBELáEZ P,GIRSHICK R,et al.Simultaneous detection and segmentation[C]//Proceedings of European Conference on Computer Vision.Switzerland:Springer International Publishing,2014:297-312.
- [8]SERMANET P,EIGEN D,ZHANG X,et al.Over Feat:integrated recognition,localization and detection using convolutional networks[J/OL].[2014-02-24].https://arxiv.org/pdf/1312.6229v4.pdf.
- [9]KRIZHEVSKY A,SUTSKEVER I,HINTON G.Image Net classification with deep convolutional neural networks[C]//Proceedings of 25th International Conference on Neural Information Processing Systems.Lake Tahoe:Curran Associates Inc.,2012:1097-1105.
- [10]HE K,ZHANG X,REN S,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J/OL].[2015-04-23].https://arxiv.org/abs/1406.4729.
- [11]WANG X,YANG M,ZHU S,et al.Regionlets for generic object detection[C]//Proceedings of IEEE International Conference on Computer Vision.Sydney:IEEE,2014:17-24.
- [12]JESORSKY O,KIRCHBERG K J,FRISCHHOLZ R W.Robust face detection using the Hausdorff distance[C]//Proceedings of 3rd International Conference on Audio-and Videobased Biometric Person Authentication.Halmstad:Springer,2001:90-95.