基于分类外形搜索的人脸特征点定位
|
摘要
针对传统由粗糙到精准的人脸外形搜索方法,其每一次外形搜索需要在整个外形搜索空间进行,提出一种基于分类的外形搜索方法。该方法始于一个包含不同人脸形状的外形搜索空间,首先利用基于相关性的特征选择方法对随机森林分类器进行优化,利用训练的随机森林分类器将外形搜索空间分为若干个外形搜索子空间;然后根据输入样本和随机森林分类器确定与当前外形最接近的外形搜索子空间,并计算对应子空间的中心和对应样本的后验概率分布,方便后续阶段更好地进行外形搜索;最后采用级联回归进行人脸特征点定位。在300-W数据集上的实验结果表明,该方法不仅有效降低了外形搜索的时间,同时在无约束环境中具有良好的鲁棒性。
Aiming at the traditional shape searching which needed to search in the whole shape space each time,this paper proposed a new approach based on classified face shape searching. This approach began with a shape space that contained diverse shapes. First it optimized the random forest classifiers by the feature selection method based on correlation and trained the random forest classifiers by training samples. Then it divided the shape space into several sub-spaces by random forest classifiers,and searched the sub-space that was the most similar to the current shape,and estimated the center of the sub-space and probability distribution. Finally,this paper employed cascaded regression to realize face alignment. The experiment result demonstrates the proposed approach obvious decreases the searching time and has good robustness in unconstrained environment on the 300-W database.
Aiming at the traditional shape searching which needed to search in the whole shape space each time,this paper proposed a new approach based on classified face shape searching. This approach began with a shape space that contained diverse shapes. First it optimized the random forest classifiers by the feature selection method based on correlation and trained the random forest classifiers by training samples. Then it divided the shape space into several sub-spaces by random forest classifiers,and searched the sub-space that was the most similar to the current shape,and estimated the center of the sub-space and probability distribution. Finally,this paper employed cascaded regression to realize face alignment. The experiment result demonstrates the proposed approach obvious decreases the searching time and has good robustness in unconstrained environment on the 300-W database.
引文
[1]Lawrence S,Giles C L,Tsoi A C,et al.Face recognition:a convolutional neural-network approach[J].IEEE Trans on Neural Networks,1997,8(1):98-113.
[2]Oneto L,Ghio A,Ridella S,et al.Out-of-sample error estimation:the blessing of high dimensionality[C]//Proc of IEEE International Conference on Data Mining.Washington DC:IEEE Computer Society,2015:637-644.
[3]Burgos-Artizzu X P,Perona P,Dollar P.Robust face landmark estimation under occlusion[C]//Proc of IEEE International Conference on Computer Vision.Washington DC:IEEE Computer Society,2013:1513-1520.
[4]Kaburlasos V G,Papadakis S E,Papakostas G A.Lattice computing extension of the FAM neural classifier for human facial expression recognition[J].IEEE Trans on Neural Networks&Learning Systems,2013,24(10):1526-1538.
[5]Weng Yanlin,Cao Chen,Hou Qiming,et al.Real-time facial animation on mobile devices[J].Graphical Models,2014,76(3):172-179.
[6]Cao Chen,Bradley D,Zhou Kun,et al.Real-time high-fidelity facial performance capture[J].ACM Trans on Graphics,2015,34(4):article No 46.
[7]Zhang Jie,Shan Shiguang,Kan Meina,et al.Coarse-to-fine auto-encoder networks(CFAN)for real-time face alignment[C]//Proc of European Conference on Computer Vision.Cham:Springer,2014:1-16.
[8]Li Quefeng,Cheng Guang,Fan Jianqing,et al.Embracing the blessing of dimensionality in factor models[J].Journal of the American Statistical Association,2018,113(521):380-389.
[9]Asthana A,Zafeiriou S,Cheng Shiyang,et al.Robust discriminative response map fitting with constrained local models[C]//Proc of IEEEConference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2013:3444-3451.
[10]Tzimiropoulos G,Pantic M.Optimization problems for fast AAM fitting in-the-wild[C]//Proc of IEEE International Conference on Computer Vision.Washington DC:IEEE Computer Society,2013:593-600.
[11]Sun Jian,Wen Fang,Wei Yichen,et al.Face alignment by explicit shape regression[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2012:2887-2894.
[12]Zhu Shizhan,Li Cheng,Loy C C,et al.Face alignment by coarse-tofine shape searching[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2015:4998-5006.
[13]Pavlidis T.In memoriam:King-Sun Fu[J].IEEE Trans on Pattern Analysis&Machine Intelligence,1985,7(4):373.
[14]Ozuysal M,Calonder M,Lepetit V,et al.Fast keypoint recognition using random ferns[J].IEEE Trans on Pattern Analysis&Machine Intelligence,2010,32(3):448-461.
[15]Liu Ruixu,Shen Ju,Sun Qingquan,et al.Cascaded pose regression revisited:face alignment in videos[C]//Proc of the 3rd IEEE International Conference on Multimedia Big Data.Washington DC:IEEEComputer Society,2017:291-298.
[16]Dosse M B,Kiers H A L,Berge J M F T.Anisotropic generalized procrustes analysis[J].Computational Statistics&Data Analysis,2011,55(5):1961-1968.
[17]Yu Le,Liu Huan.Efficient feature selection via analysis of relevance and redundancy[J].Journal of Machine Learning Research,2004,5(12):1205-1224.
[18]蒋盛益,王连喜.基于相关性的特征选择[J].计算机工程与应用,2010,46(20):153-156.(Jiang Shengyi,Wang Lianxi.Feature selection based on feature similarity measure[J].Computer Engineering and Application,2010,46(20):153-156.)
[19]University of California-Irvine.UCI repository of machine learning database[DB/OL].(1998).http://www.ics.uci.edu/?mlearn/ML-Repository.html.
[20]Sagonas C,Tzimiropoulos G,Zafeiriou S,et al.300 faces in-the-wild challenge:the first facial landmark localization challenge[C]//Proc of IEEE International Conference on Computer Vision.Washington DC:IEEE Computer Society,2013:397-403.
[21]Ren Shaoqing,Cao Xudong,Wei Yichen,et al.Face alignment at3000 FPS via regressing local binary features[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2014:1685-1692.
[22]Xiong Xuehan,De la Torre F.Supervised descent method and its applications to face alignment[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2013:532-539.
[2]Oneto L,Ghio A,Ridella S,et al.Out-of-sample error estimation:the blessing of high dimensionality[C]//Proc of IEEE International Conference on Data Mining.Washington DC:IEEE Computer Society,2015:637-644.
[3]Burgos-Artizzu X P,Perona P,Dollar P.Robust face landmark estimation under occlusion[C]//Proc of IEEE International Conference on Computer Vision.Washington DC:IEEE Computer Society,2013:1513-1520.
[4]Kaburlasos V G,Papadakis S E,Papakostas G A.Lattice computing extension of the FAM neural classifier for human facial expression recognition[J].IEEE Trans on Neural Networks&Learning Systems,2013,24(10):1526-1538.
[5]Weng Yanlin,Cao Chen,Hou Qiming,et al.Real-time facial animation on mobile devices[J].Graphical Models,2014,76(3):172-179.
[6]Cao Chen,Bradley D,Zhou Kun,et al.Real-time high-fidelity facial performance capture[J].ACM Trans on Graphics,2015,34(4):article No 46.
[7]Zhang Jie,Shan Shiguang,Kan Meina,et al.Coarse-to-fine auto-encoder networks(CFAN)for real-time face alignment[C]//Proc of European Conference on Computer Vision.Cham:Springer,2014:1-16.
[8]Li Quefeng,Cheng Guang,Fan Jianqing,et al.Embracing the blessing of dimensionality in factor models[J].Journal of the American Statistical Association,2018,113(521):380-389.
[9]Asthana A,Zafeiriou S,Cheng Shiyang,et al.Robust discriminative response map fitting with constrained local models[C]//Proc of IEEEConference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2013:3444-3451.
[10]Tzimiropoulos G,Pantic M.Optimization problems for fast AAM fitting in-the-wild[C]//Proc of IEEE International Conference on Computer Vision.Washington DC:IEEE Computer Society,2013:593-600.
[11]Sun Jian,Wen Fang,Wei Yichen,et al.Face alignment by explicit shape regression[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2012:2887-2894.
[12]Zhu Shizhan,Li Cheng,Loy C C,et al.Face alignment by coarse-tofine shape searching[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2015:4998-5006.
[13]Pavlidis T.In memoriam:King-Sun Fu[J].IEEE Trans on Pattern Analysis&Machine Intelligence,1985,7(4):373.
[14]Ozuysal M,Calonder M,Lepetit V,et al.Fast keypoint recognition using random ferns[J].IEEE Trans on Pattern Analysis&Machine Intelligence,2010,32(3):448-461.
[15]Liu Ruixu,Shen Ju,Sun Qingquan,et al.Cascaded pose regression revisited:face alignment in videos[C]//Proc of the 3rd IEEE International Conference on Multimedia Big Data.Washington DC:IEEEComputer Society,2017:291-298.
[16]Dosse M B,Kiers H A L,Berge J M F T.Anisotropic generalized procrustes analysis[J].Computational Statistics&Data Analysis,2011,55(5):1961-1968.
[17]Yu Le,Liu Huan.Efficient feature selection via analysis of relevance and redundancy[J].Journal of Machine Learning Research,2004,5(12):1205-1224.
[18]蒋盛益,王连喜.基于相关性的特征选择[J].计算机工程与应用,2010,46(20):153-156.(Jiang Shengyi,Wang Lianxi.Feature selection based on feature similarity measure[J].Computer Engineering and Application,2010,46(20):153-156.)
[19]University of California-Irvine.UCI repository of machine learning database[DB/OL].(1998).http://www.ics.uci.edu/?mlearn/ML-Repository.html.
[20]Sagonas C,Tzimiropoulos G,Zafeiriou S,et al.300 faces in-the-wild challenge:the first facial landmark localization challenge[C]//Proc of IEEE International Conference on Computer Vision.Washington DC:IEEE Computer Society,2013:397-403.
[21]Ren Shaoqing,Cao Xudong,Wei Yichen,et al.Face alignment at3000 FPS via regressing local binary features[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2014:1685-1692.
[22]Xiong Xuehan,De la Torre F.Supervised descent method and its applications to face alignment[C]//Proc of IEEE Conference on Computer Vision and Pattern Recognition.Washington DC:IEEE Computer Society,2013:532-539.