基于代价敏感深度决策树的公交车环境人脸检测

doi:10.11772/j.issn.1001-9081.2017.11.3152

计算机应用 ›› 2017, Vol. 37 ›› Issue (11): 3152-3156.DOI: 10.11772/j.issn.1001-9081.2017.11.3152

• 第十六届中国机器学习会议(CCML 2017) • 上一篇下一篇

基于代价敏感深度决策树的公交车环境人脸检测

娄康^1,2, 薛彦兵^1,2, 张桦^1,2, 徐光平^1,2, 高赞^1,2, 王志岗^1,2

1. 计算机视觉与系统省部共建教育部重点实验室(天津理工大学), 天津 300384;
2. 天津市智能计算及软件新技术重点实验室(天津理工大学), 天津 300384

收稿日期:2017-05-16 修回日期:2017-06-05 出版日期:2017-11-10 发布日期:2017-11-11
通讯作者: 薛彦兵
作者简介:娄康(1992-),女,山东聊城人,硕士研究生,CCF会员,主要研究方向:计算机视觉,机器学习;薛彦兵(1979-),男,山东沂南人,副研究员,硕士,CCF会员,主要研究方向:计算机视觉、机器学习;张桦(1962-),女,四川安丘人,教授,博士生导师,博士,CCF会员,主要研究方向:计算机视觉、机器学习。
基金资助:
国家自然科学基金资助项目（U1509207，61325019，61472278，61403281，61572357）。

Face detection in bus environment based on cost-sensitive deep quadratic tree

LOU Kang^1,2, XUE Yanbing^1,2, ZHANG Hua^1,2, XU Guangping^1,2, GAO Zan^1,2, WANG Zhigang^1,2

1. Key Laboratory of Computer Vision and System of Ministry of Education(Tianjin University of Technology), Tianjin 300384, China;
2. Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology(Tianjin University of Technology), Tianjin 300384, China

Received:2017-05-16 Revised:2017-06-05 Online:2017-11-10 Published:2017-11-11
Supported by:
This work is partially supported by the National Natural Foundation of China (U1509207,61325019,61472278, 61403281, 61572357).

摘要/Abstract

摘要： 针对公交车环境下的人脸检测具有光照变化、模糊、遮挡、低分辨率和姿势变化等问题，提出了基于代价敏感深度决策树的人脸检测算法。首先，基于归一化的像素差异（NPD）特征构建单个深度二次树（DQT）；接着，根据当前决策树的分类结果，利用代价敏感Gentle Adaboost方法对样本权重进行更新，依次训练出多棵深度决策树；最后，将所有决策树通过Soft-Cascade级联得到最终的检测算法。在人脸检测数据集（FDDB）和公交车视频上的实验结果表明，所提算法与现有的深度决策树算法相比，在检测率和检测速度上均有提升。

关键词: 归一化的像素差异特征, 代价敏感, 深度二次树, Gentle Adaboost方法, Soft-Cascade

Abstract: The problems of face detection in bus environment include ambient illumination changing, image distortion, human body occlusion, abnormal postures and etc. For alleviating these mentioned limitations, a face detection based on cost-sensitive Deep Quadratic Tree (DQT) was proposed. First of all, Normalized Pixel Difference (NPD) feature was utilized to construct and train a single DQT. According to the classification result of the current decision tree, the cost-sensitive Gentle Adaboost method was used to update the sample weight, and a number of deep decision trees were trained. Finally, the classifier was produced by Soft-Cascade method with multiple upgraded deep quadratic trees. The experimental results on Face Detection Data set and Benchmark (FDDB) and bus video show that compared with the existing depth decision tree algorithm, the proposed algorithm has improved the detection rate and detection speed.

Key words: Normalized Pixel Difference (NPD) feature, cost-sensitive, Deep Quadratic Tree (DQT), Gentle Adaboost method, Soft-Cascade

中图分类号:

TP391.41

娄康, 薛彦兵, 张桦, 徐光平, 高赞, 王志岗. 基于代价敏感深度决策树的公交车环境人脸检测[J]. 计算机应用, 2017, 37(11): 3152-3156.

LOU Kang, XUE Yanbing, ZHANG Hua, XU Guangping, GAO Zan, WANG Zhigang. Face detection in bus environment based on cost-sensitive deep quadratic tree[J]. Journal of Computer Applications, 2017, 37(11): 3152-3156.

参考文献

[1] VIOLA P, JONES M. Rapid object detection using a boosted cascade of simple features[C]//Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2003:511-518.
[2] LIENHART R, MAYDT J. An extended set of Haar-like features for rapid object detection[C]//Proceedings of the 2002 International Conference on Image Processing. Piscataway, NJ:IEEE, 2002:900-903.
[3] LI S, ZHU L, ZHANG Z, et al. Statistical learning of multi-view face detection[C]//Proceedings of the 7th European Conference on Computer Vision. London:Springer-Verlag, 2002:67-81.
[4] JONES M, VIOLA P. Fast multi-view face detection[C]//Proceedings of the 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2003:276-286
[5] WU B, AI H, HUANG C et al. Fast rotation invariant multi-view face detection based on real AdaBoost[C]//Proceedings of the 6th IEEE International Conference on Automatic Face and Gesture Recognition. Washington, DC:IEEE Computer Society, 2004:79-84.
[6] FROBA B, ERNST A. Face detection with the modified census transform[C]//Proceedings of the 6th IEEE International Conference on Automatic Face and Gesture Recognition. Washington, DC:IEEE Computer Society, 2004:91-96.
[7] JIN H, LIU Q, LU H, et al. Face detection using improved LBP under Bayesian framework[C]//Proceedings of the 2004 IEEE First Symposium on Multi-Agent Security and Survivability. Washington, DC:IEEE Computer Society, 2004:306-309.
[8] MITA T, KANEKO T, HORI O. Joint Haar-like features for face detection[C]//Proceedings of the 10th IEEE International Conference on Computer Vision. Washington, DC:IEEE Computer Society, 2005:1619-1626.
[9] ZHANG H, GAO W, CHEN X, et al. Object detection using spatial histogram features[J]. Image and Vision Computing, 2006, 24(4):327-341.
[10] ZHANG L, CHU R, XIANG S, et al. Face detection based on multi-block LBP representation[C]//Proceedings of the 2007 International Conference on Advances in Biometrics. Berlin:Springer-Verlag, 2007:11-18.
[11] LI S, ZHANG Z. FloatBoost learning and statistical face detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(9):11-12.
[12] HUANG C, AI H, LI Y, et al. High-performance rotation invariant multiview face detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(4):671-686.
[13] FELZENSZWALB P F, GIRSHICK R B, MCALLESTER D, et al. Object detection with discriminatively trained part-based models[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(9):1627-1645.
[14] MARKUŠ N, FRLJAK M, PANDŽIC I S, et al. Object detection with pixel intensity comparisons organized in decision trees[J]. Computer Science, 2013, 14(4):2657-2662.
[15] LIAO S, JAIN A K, LI S Z.A. Fast and accurate unconstrained face detector[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2016, 38(2):211.
[16] 李航.统计学习方法[M].北京:清华大学出版社, 2012:55-75.(LI H. Statistical Learning Methods[M]. Beijing:Tsinghua University Press, 2012:55-75.)
[17] 李秋洁, 茅耀斌, 叶曙光,等.代价敏感Boosting算法研究[J].南京理工大学学报(自然科学版), 2013, 37(1):23-28.(LI Q J, MAO Y B, YE S G, et al. Cost-sensitive Boosting algorithm[J].Journal of Nanjing University of Science and Technology, 2013, 37(1):23-28.)
[18] 曹二奎.基于Gentle Adaboost的人脸检测算法研究[D]. 西安:西安电子科技大学, 2013:49-60.(CAO E K. Research on face detection algorithm based on Gentle Adaboost[D]. Xi'an:Xidian University, 2013:49-60.)
[19] 曹莹, 苗启广, 刘家辰,等. 具有Fisher一致性的代价敏感Boosting算法[J].软件学报, 2013,24(11):2584-2596.(CAO Y, MIAO Q G, LIU J C, et al. Fisher consistent cost sensitive Boosting algorithm[J].Journal of Software, 2013,24(11):2584-2596.)
[20] BOURDEV L, BRANDT J. Robust object detection via soft cascade[C]//Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Washington, DC:IEEE Computer Society, 2005:236-243.
[21] KÖSTINGER M, WOHLHART P, ROTH P M, et al. Annotated facial landmarks in the wild:a large-scale, real-world database for facial landmark localization[C]//Proceedings of the 2011 IEEE International Conference on Computer Vision Workshops. Piscataway, NJ:IEEE, 2011:2144-2151.
[22] JAIN V, LEARNED-MILLER E. FDDB:a benchmark for face detection in unconstrained settings[EB/OL].[2016-11-20]. http://www.cics.umass.edu/~elm/papers/fddb.pdf.
[23] MIKOLAJCZYK K, SCHMID C, ZISSERMAN A. Human detection based on a probabilistic assembly of robust part detectors[C]//Proceedings of the 8th European Conference on Computer Vision. Berlin:Springer-Verlag, 2004:69-82.
[24] SHEN X, LIN Z, BRANDT J et al. Detecting and aligning faces by image retrieval[C]//Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition. Washington, DC:IEEE Computer Society, 2013:3460-3467.
[25] KÖSTINGER M, WOHLHART P, ROTH P M, et al. Robust face detection by simple means[EB/OL].[2016-11-20]. http://dagm2012.icg.tugraz.at/cvaw12/procs/02_Koestinger_cvaw2012.pdf.

基于代价敏感深度决策树的公交车环境人脸检测

Face detection in bus environment based on cost-sensitive deep quadratic tree

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 10

编辑推荐

Metrics

[1]	孟凡, 陈广, 王勇, 高阳, 高德群, 贾文龙. 基于多粒度时序结构表示的异常检测算法在储层含油性检测中应用[J]. 计算机应用, 2021, 41(8): 2453-2459.
[2]	王俊红, 闫家荣. 基于欠采样和代价敏感的不平衡数据分类算法[J]. 计算机应用, 2021, 41(1): 48-52.
[3]	杨浩, 王宇, 张中原. 基于K最近邻样本平均距离的代价敏感算法的集成[J]. 计算机应用, 2019, 39(7): 1883-1887.
[4]	王莉, 陈红梅, 王生武. 新的基于代价敏感集成学习的非平衡数据集分类方法NIBoost[J]. 计算机应用, 2019, 39(3): 629-633.
[5]	郭冰楠, 吴广潮. 基于改进的代价敏感决策树的网络贷款分类[J]. 计算机应用, 2019, 39(10): 2888-2892.
[6]	谭洁帆, 朱焱, 陈同孝, 张真诚. 基于卷积神经网络和代价敏感的不平衡图像分类方法[J]. 计算机应用, 2018, 38(7): 1862-1865.
[7]	任福龙, 曹鹏, 万超, 赵大哲. 结合代价敏感半监督集成学习的糖尿病视网膜病变分级[J]. 计算机应用, 2018, 38(7): 2124-2129.
[8]	刘星毅. 一种新的分裂属性选择方法[J]. 计算机应用, 2009, 29(3): 839-842.
[9]	何海江. 代价与样本相关的简约核支持向量机[J]. 计算机应用, 2008, 28(11): 2863-2866.
[10]	罗菲菲;刘贵全;安景琦;张婷慧. 一种基于代价敏感学习的范例推理方法及其应用研究[J]. 计算机应用, 2005, 25(10): 2444-2446.