基于LPP的视频图像头部姿态估计的方法研究
|
摘要
头部姿态估计在注意力检测、行为检测和人脸识别上有着重要的研究意义。局部保持投影算法为人们处理非线性降维问题提供了一种新的思路。作为一种线性降维方法,在头部姿态估计领域具有很大的应用空间。
对于头部姿态估计问题,本文研究主要围绕有监督的局部保持投影和异常值的度量方法进行展开。鉴于无监督局部保持投影的流形学习算法对头部姿态估计的高误差性和对噪音的敏感性问题,本文设计正弦偏置距离方法和融入带权值主成分分析方法来改进局部保持投影算法。首先对训练的头部姿态加以姿态标注,并获得各个头部姿态间的正弦偏置距离;然后对所有头部姿态样本点进行异常值的度量,训练出较好的线性映射矩阵,再采用改进后的局部保持投影算法对图像进行降维处理;最后采用支持向量机分类器进行头部姿态估计。
用改进的局部保持投影方法进行了头部姿态估计实验。由于融入了正弦偏置距离和带权值主成分分析方法,不仅有效地消除人的身份的影响,还很好地削弱光照变化、表情变化、噪音等因素的影响。并且大量实验也表明:改进后的局部保持投影算法同改进前局部保持投影算法、局部嵌入分析算法相比,无论在静态的头部姿态数据库中,还是在动态的视频流中,头部姿态估计都取得较好的效果。
Head pose estimation is an important research issue in attention test, behavioral detection, and face recognition. Locality Preserving Projection method provides a fresh idea for us to do nonlinear dimensionality. As a linear dimension reduction tool, Locality Preserving Projection method has great potential to be applied in head pose estimation.
This research concentrates on Supervised Locality Preserving Projection and Outlier Measure method to start. Aiming at the problems of the high head pose estimation error and the noise sensitivity of unsupervised LPP, Sinusoidal offset distance method and weighted PCA method was used to estimate the head pose. By this algorithm, the head poses of the training samples are firstly labeled, and all the biased distances between the head poses are calculated. And then the outliers of the head samples are detected, so as to train a better linear mapping matrix, then using the Improved LPP to reduce the dimensions of the image; the SVM classifier are finally used to estimate the head pose.
In this paper, the Improved LPP method is used to estimate the head pose. Sinusoidal offset distance method and weighted PCA method not only effectively eliminate the impact of the identity, but also effectively reduce the impact of illumination, facial expression changes and noise. The head pose estimation experiments show that the improved LPP achieves the better results than the traditional LPP and LEA algorithms in both static head pose database and dynamic video stream.
对于头部姿态估计问题,本文研究主要围绕有监督的局部保持投影和异常值的度量方法进行展开。鉴于无监督局部保持投影的流形学习算法对头部姿态估计的高误差性和对噪音的敏感性问题,本文设计正弦偏置距离方法和融入带权值主成分分析方法来改进局部保持投影算法。首先对训练的头部姿态加以姿态标注,并获得各个头部姿态间的正弦偏置距离;然后对所有头部姿态样本点进行异常值的度量,训练出较好的线性映射矩阵,再采用改进后的局部保持投影算法对图像进行降维处理;最后采用支持向量机分类器进行头部姿态估计。
用改进的局部保持投影方法进行了头部姿态估计实验。由于融入了正弦偏置距离和带权值主成分分析方法,不仅有效地消除人的身份的影响,还很好地削弱光照变化、表情变化、噪音等因素的影响。并且大量实验也表明:改进后的局部保持投影算法同改进前局部保持投影算法、局部嵌入分析算法相比,无论在静态的头部姿态数据库中,还是在动态的视频流中,头部姿态估计都取得较好的效果。
Head pose estimation is an important research issue in attention test, behavioral detection, and face recognition. Locality Preserving Projection method provides a fresh idea for us to do nonlinear dimensionality. As a linear dimension reduction tool, Locality Preserving Projection method has great potential to be applied in head pose estimation.
This research concentrates on Supervised Locality Preserving Projection and Outlier Measure method to start. Aiming at the problems of the high head pose estimation error and the noise sensitivity of unsupervised LPP, Sinusoidal offset distance method and weighted PCA method was used to estimate the head pose. By this algorithm, the head poses of the training samples are firstly labeled, and all the biased distances between the head poses are calculated. And then the outliers of the head samples are detected, so as to train a better linear mapping matrix, then using the Improved LPP to reduce the dimensions of the image; the SVM classifier are finally used to estimate the head pose.
In this paper, the Improved LPP method is used to estimate the head pose. Sinusoidal offset distance method and weighted PCA method not only effectively eliminate the impact of the identity, but also effectively reduce the impact of illumination, facial expression changes and noise. The head pose estimation experiments show that the improved LPP achieves the better results than the traditional LPP and LEA algorithms in both static head pose database and dynamic video stream.
引文
[1] P. J. Phillips, P. Grother, R. J. Michaels, et al. Face Recognition Vendor Test 2002: Evaluation Report[R]. National Institute of Standards and Technology, USA. 2003.
[2] V.F. Ferrario, C. Sforza G. Serrao, G. Grassi, and E. Mossi,“Active Range of Motion of the Head and Cervical Spine: A Three-Dimensional Investigation in Healthy Young Adults[J]. J. Orthopaedic Research, vol. 20, no. 1, pp. 122-129, 2002.
[3] J. Ng and S. Gong, Composite Support Vector Machines for Detection of Faces Across Views and Pose Estimation[J]. Image and Vision Computing, vol. 20, nos. 5-6, pp. 359-368, 2002.
[4] J. Ng and S. Gong, Multi-View Face Detection and Pose Estimation Using a Composite Support Vector Machine Across the View Sphere[J]. Proc. IEEE Int’l Workshop Recognition, Analysis, and Tracking of Faces and Gestures in Real-Time Systems, pp. 14-21,1999.
[5] H. Rowley, S. Baluja, and T. Kanade, Rotation Invariant Neural Network-Based Face Detection[J]. Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 38-44, 1998.
[6] M. Jones and P. Viola, Fast Multi-View Face Detection[R]. Technical Report 096, Mitsubishi Electric Research Laboratories,2003.
[7] A. Gee and R. Cipolla. Fast Visual Tracking by TemporalConsensus[J]. Image and Vision Computing, vol. 14, no. 2, pp. 105-114, 1996.
[8]T. Cootes, K. Walker, and C. Taylor, View-Based Active Appearance Models[J]. Proc. IEEE Int’l Conf. Automatic Face and Gesture Recognition, pp. 227-232, 2000.
[9] T. Cootes, G. Edwards, and C. Taylor, Active Appearance Models[J]. IEEE Trans. Pattern Analysis and Machine Intelligence,vol. 23, no. 6, pp. 681-685, June 2001.
[10] M. Cordea, E. Petriu, N. Georganas, D. Petriu, and T. Whalen, Real-Time 2(1/2)-D Head Pose Recovery for Model-Based Video-Coding[J]. IEEE Trans. Instrumentation and Measurement, vol. 50,no. 4, pp. 1007-1013, 2001.
[11] R. Duda, P. Hart, and D. Stork, Pattern Classification[J]. second ed. John Wiley& Sons, 2001.
[12] S. Roweis and L. Saul, Nonlinear Dimensionality Reduction by Locally Linear Embedding[J]. Science, vol. 290, no. 5500, pp. 2323-2326, 2000.
[13] M. Belkin and P. Niyogi, Laplacian Eigenmaps for Dimensionality Reduction and Data Representation[J]. Neural Computation, vol. 15, no. 6, pp. 1373-1396, 2003.
[14] B. Raytchev, I. Yoda, and K. Sakaue, Head Pose Estimation by Nonlinear Manifold Learning[J]. Proc. 17th Int’l Conf. PatternRecognition, pp. 462-466, 2004.
[15] Y. Fu and T. Huang, Graph Embedded Analysis for Head Pose Estimation[J]. Proc. IEEE Int’l Conf. Automatic Face and Gesture Recognition, pp. 3-8, 2006.
[16] X.He,Partha Niyogi,”Locality Preserving Projections,“2005
[17] Z. Li, Y. Fu, J. Yuan, T. Huang, and Y. Wu, Query Driven Localized Linear Discriminant Models for Head Pose Estimation[J]. Proc. IEEE Int’l Conf. Multimedia and Expo, pp. 1810-1813, 2007.
[18]翁仲毅,萧志杰,局部保持投影算法[J],2003.
[19] S. Roweis and L. Saul, Nonlinear Dimensionality Reduction by Locally Linear Embedding[J]. Science, vol. 290, no. 5500, pp. 2323-2326, 2000.
[20] Xianhua Zeng, Siwei Luo, Generalized Locally Linear Embedding Based on Local Reconstruction Similarity[J]. 2008 Fifth International Conference on Fuzzy Systems and Knowledge Discovery, 2008, vol. 5, pp.305-309.
[21] Z. Lu, M. C. Perpinan, and C. Sminchisescu. People Tracking with the Laplacian Eigenmaps Latent Variable Model[J]. Advances in Neural Information Processing Systems, 2007. Volume: 20, Pages: 1705-1712.
[22] Weiwei Yu, Face recognition using discriminant locality preserving projections[J]. Image and Vision Computing Volume 24, Issue 3, 1 March 2006, Pages 239-248.
[23] Hai-liang Feng, Research on Applying Manifold Learning Algorithms to Face Recognition[D].College of Optoelectronic Engineering of Chongqing University,2008. (冯海亮.流形学习算法在人脸识别中的应用研究[D].重庆大学光电工程学院,2008)
[24] LI Zheng-yi, ZHU Yi-dan, Face Recognition Based on Supervised DirectLocality Preserving Projection[J]. Computer Engineering, Vol.35 No.10,Panges:190-192. (李政仪,朱益丹,基于有监督直接局部保持投影的人脸识别[J].计算机工程,2009,35(10),起止页数:190-192).
[25] Z. Guo, H. Liu, Q. Wang, and J. Yang, A Fast Algorithm Face Detection and Head Pose Estimation for Driver Assistant System[J].Proc. Eighth Int’l Conf. Signal Processing, 2006, Page(s): 1165– 1570 Vol.3.
[26] N. Hu, W. Huang, and S. Ranganath, Head Pose Estimation by Non-Linear Embedding and Mapping[D]. Proc. IEEE Int’l Conf.Image Processing, vol. 2, pp. 342-345, 2005.
[27] Erik Murphy-Chutorian, Mohan Manubhai Trivedi, Head Pose Estimation in Computer Vision: A Survey[J]. IEEE transactions on pattern analysis and machine intelligence, 2009,USA ,607– 626.
[28]边肇祺,张学共著.模式识别[M].北京:清华大学出版社,2007.
[29] Areans Garcia J,Perez Gruz F. Multi-classes Support Vector Machine: A New Approach[A]. Proc of the IEEE Int Conf on A coustics, Speech and Signal Processing[C]. New York 2003, 2:781-784
[30]余晖,赵晖.支持向量机多类分类算法新研究[J].计算机工程与应用,2008,44(7):185-189.
[31] B.Raytchev, I. Yoda and K.Sakaue, Head Pose Estimation By Nonlinear Manifold Learning[J]. IEEE Conf.on ICPR’04, vol. 4, pp. 462-466 , 2004.
[32] Z. Li, Y. Fu,J. Yuan, T. Huang. Query Driven Localized Linear Discriminant Models for Head Pose Estimation[J] . Proc.IEEE Int’l Conf. Multimedia and Expo,pp.1810-1813, 2007.
[33] Hong Qiao, Peng Zhang, Bo Zhang, Suiwu Zheng ,Learning an Intrinsic Variable Preserving Manifold for Dynamic Visual Tracking[J]. IEEE Trans Syst Man Cybern B Cybern. 2010 Jun; Volume: 40 Issue:3, On page(s): 868– 880.
[34] V. Balasubramanian, J. Ye, and S.Panchanathan, Biased Manifold Embedding: A Frameworkfor Person Independent Head Pose Estimation [J].Proc. IEEE Conf .Computer Vision and Pattern Recognition, 2007. 1-4244-1180-7 , Page(s): 1– 7.
[35] LI Rui-fan,ZHU Qiang-sheng,Rubust Locality Preserving Projection for Facial Expression Recognition[J].Journal of Beijing University of Posts and Telecommunications,2006, Volume:29(Z2),Pages:178-182. (李睿凡,朱强生,鲁棒局部保持投影的表情识别[J],北京邮电大学学报,2006,29(Z2) 178-182) .
[36] Little,D.Krishna,S,Black,J.Panchanathan,S.A Methodlogy for Evaluating Robustness of Face Recognition Algorithms with Respect to Variations in Pose angle and Illumination Angle[J]. IEEE International Conference on Acoustics,Speech, and Signal Processing,2005.
[37] Pointing’04 DataBase: http://www-prima.inrialpes.fr/Pointing’04.
[38] Vineeth Nallure Sreekar Krishna, Sethuraman PanchanathanPerson-Independent. Head Pose Estimation UsingBiased Manifold Embedding[J]. EURASIP Journal on Advances in Signal Processing Volume 2008, Article ID 283540, 15 pages.
[39]张旭光,赵恩良,王延杰.基于Mean-Shift的灰度目标跟踪新算法[J].光学技术, Vol. 33,pp. 226-229, 2007.
[40] Cheng, Y. Mean-shift, Mode Seeking and Clustering[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 17, pp. 790, 1995.
[41] Comaniciu D, Ramesh V, Meer P. Kernel-based Object Tracking[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 25, pp. 564-577, 2003.
[42] Djouadi A,Snorrason 0,Garber F D. The Quality of Training-Sample Estimates of the Bhattacharyya Coefficient[J].IEEE Trans on Pattern Analysis and Machine Intelligent,1990,12:92—97.
[43] Nummiaro K, Koller-Meier E and Gool L.V. An Adaptive Color-based Particle Filter[J]. Image and Vision Computing, Vol. 21, pp. 99–110, 2003.
[44] Yang C, Duraiswami R and Davis L. Fast Multiple Object Tracking via a Hierarchical Particle Filter[J]. In The Proceedings of the Tenth IEEE International Conference on Computer Vision, (ICCV 2005), Beijing, China, pp. 212–219, 2005.
[45] Joachims Y. Estimating the Generalization Performance of a SVM Efficiently[A], in Proceedings of the International Conference on Machine Lerning[C], Morgan Kaufman,2000
[2] V.F. Ferrario, C. Sforza G. Serrao, G. Grassi, and E. Mossi,“Active Range of Motion of the Head and Cervical Spine: A Three-Dimensional Investigation in Healthy Young Adults[J]. J. Orthopaedic Research, vol. 20, no. 1, pp. 122-129, 2002.
[3] J. Ng and S. Gong, Composite Support Vector Machines for Detection of Faces Across Views and Pose Estimation[J]. Image and Vision Computing, vol. 20, nos. 5-6, pp. 359-368, 2002.
[4] J. Ng and S. Gong, Multi-View Face Detection and Pose Estimation Using a Composite Support Vector Machine Across the View Sphere[J]. Proc. IEEE Int’l Workshop Recognition, Analysis, and Tracking of Faces and Gestures in Real-Time Systems, pp. 14-21,1999.
[5] H. Rowley, S. Baluja, and T. Kanade, Rotation Invariant Neural Network-Based Face Detection[J]. Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 38-44, 1998.
[6] M. Jones and P. Viola, Fast Multi-View Face Detection[R]. Technical Report 096, Mitsubishi Electric Research Laboratories,2003.
[7] A. Gee and R. Cipolla. Fast Visual Tracking by TemporalConsensus[J]. Image and Vision Computing, vol. 14, no. 2, pp. 105-114, 1996.
[8]T. Cootes, K. Walker, and C. Taylor, View-Based Active Appearance Models[J]. Proc. IEEE Int’l Conf. Automatic Face and Gesture Recognition, pp. 227-232, 2000.
[9] T. Cootes, G. Edwards, and C. Taylor, Active Appearance Models[J]. IEEE Trans. Pattern Analysis and Machine Intelligence,vol. 23, no. 6, pp. 681-685, June 2001.
[10] M. Cordea, E. Petriu, N. Georganas, D. Petriu, and T. Whalen, Real-Time 2(1/2)-D Head Pose Recovery for Model-Based Video-Coding[J]. IEEE Trans. Instrumentation and Measurement, vol. 50,no. 4, pp. 1007-1013, 2001.
[11] R. Duda, P. Hart, and D. Stork, Pattern Classification[J]. second ed. John Wiley& Sons, 2001.
[12] S. Roweis and L. Saul, Nonlinear Dimensionality Reduction by Locally Linear Embedding[J]. Science, vol. 290, no. 5500, pp. 2323-2326, 2000.
[13] M. Belkin and P. Niyogi, Laplacian Eigenmaps for Dimensionality Reduction and Data Representation[J]. Neural Computation, vol. 15, no. 6, pp. 1373-1396, 2003.
[14] B. Raytchev, I. Yoda, and K. Sakaue, Head Pose Estimation by Nonlinear Manifold Learning[J]. Proc. 17th Int’l Conf. PatternRecognition, pp. 462-466, 2004.
[15] Y. Fu and T. Huang, Graph Embedded Analysis for Head Pose Estimation[J]. Proc. IEEE Int’l Conf. Automatic Face and Gesture Recognition, pp. 3-8, 2006.
[16] X.He,Partha Niyogi,”Locality Preserving Projections,“2005
[17] Z. Li, Y. Fu, J. Yuan, T. Huang, and Y. Wu, Query Driven Localized Linear Discriminant Models for Head Pose Estimation[J]. Proc. IEEE Int’l Conf. Multimedia and Expo, pp. 1810-1813, 2007.
[18]翁仲毅,萧志杰,局部保持投影算法[J],2003.
[19] S. Roweis and L. Saul, Nonlinear Dimensionality Reduction by Locally Linear Embedding[J]. Science, vol. 290, no. 5500, pp. 2323-2326, 2000.
[20] Xianhua Zeng, Siwei Luo, Generalized Locally Linear Embedding Based on Local Reconstruction Similarity[J]. 2008 Fifth International Conference on Fuzzy Systems and Knowledge Discovery, 2008, vol. 5, pp.305-309.
[21] Z. Lu, M. C. Perpinan, and C. Sminchisescu. People Tracking with the Laplacian Eigenmaps Latent Variable Model[J]. Advances in Neural Information Processing Systems, 2007. Volume: 20, Pages: 1705-1712.
[22] Weiwei Yu, Face recognition using discriminant locality preserving projections[J]. Image and Vision Computing Volume 24, Issue 3, 1 March 2006, Pages 239-248.
[23] Hai-liang Feng, Research on Applying Manifold Learning Algorithms to Face Recognition[D].College of Optoelectronic Engineering of Chongqing University,2008. (冯海亮.流形学习算法在人脸识别中的应用研究[D].重庆大学光电工程学院,2008)
[24] LI Zheng-yi, ZHU Yi-dan, Face Recognition Based on Supervised DirectLocality Preserving Projection[J]. Computer Engineering, Vol.35 No.10,Panges:190-192. (李政仪,朱益丹,基于有监督直接局部保持投影的人脸识别[J].计算机工程,2009,35(10),起止页数:190-192).
[25] Z. Guo, H. Liu, Q. Wang, and J. Yang, A Fast Algorithm Face Detection and Head Pose Estimation for Driver Assistant System[J].Proc. Eighth Int’l Conf. Signal Processing, 2006, Page(s): 1165– 1570 Vol.3.
[26] N. Hu, W. Huang, and S. Ranganath, Head Pose Estimation by Non-Linear Embedding and Mapping[D]. Proc. IEEE Int’l Conf.Image Processing, vol. 2, pp. 342-345, 2005.
[27] Erik Murphy-Chutorian, Mohan Manubhai Trivedi, Head Pose Estimation in Computer Vision: A Survey[J]. IEEE transactions on pattern analysis and machine intelligence, 2009,USA ,607– 626.
[28]边肇祺,张学共著.模式识别[M].北京:清华大学出版社,2007.
[29] Areans Garcia J,Perez Gruz F. Multi-classes Support Vector Machine: A New Approach[A]. Proc of the IEEE Int Conf on A coustics, Speech and Signal Processing[C]. New York 2003, 2:781-784
[30]余晖,赵晖.支持向量机多类分类算法新研究[J].计算机工程与应用,2008,44(7):185-189.
[31] B.Raytchev, I. Yoda and K.Sakaue, Head Pose Estimation By Nonlinear Manifold Learning[J]. IEEE Conf.on ICPR’04, vol. 4, pp. 462-466 , 2004.
[32] Z. Li, Y. Fu,J. Yuan, T. Huang. Query Driven Localized Linear Discriminant Models for Head Pose Estimation[J] . Proc.IEEE Int’l Conf. Multimedia and Expo,pp.1810-1813, 2007.
[33] Hong Qiao, Peng Zhang, Bo Zhang, Suiwu Zheng ,Learning an Intrinsic Variable Preserving Manifold for Dynamic Visual Tracking[J]. IEEE Trans Syst Man Cybern B Cybern. 2010 Jun; Volume: 40 Issue:3, On page(s): 868– 880.
[34] V. Balasubramanian, J. Ye, and S.Panchanathan, Biased Manifold Embedding: A Frameworkfor Person Independent Head Pose Estimation [J].Proc. IEEE Conf .Computer Vision and Pattern Recognition, 2007. 1-4244-1180-7 , Page(s): 1– 7.
[35] LI Rui-fan,ZHU Qiang-sheng,Rubust Locality Preserving Projection for Facial Expression Recognition[J].Journal of Beijing University of Posts and Telecommunications,2006, Volume:29(Z2),Pages:178-182. (李睿凡,朱强生,鲁棒局部保持投影的表情识别[J],北京邮电大学学报,2006,29(Z2) 178-182) .
[36] Little,D.Krishna,S,Black,J.Panchanathan,S.A Methodlogy for Evaluating Robustness of Face Recognition Algorithms with Respect to Variations in Pose angle and Illumination Angle[J]. IEEE International Conference on Acoustics,Speech, and Signal Processing,2005.
[37] Pointing’04 DataBase: http://www-prima.inrialpes.fr/Pointing’04.
[38] Vineeth Nallure Sreekar Krishna, Sethuraman PanchanathanPerson-Independent. Head Pose Estimation UsingBiased Manifold Embedding[J]. EURASIP Journal on Advances in Signal Processing Volume 2008, Article ID 283540, 15 pages.
[39]张旭光,赵恩良,王延杰.基于Mean-Shift的灰度目标跟踪新算法[J].光学技术, Vol. 33,pp. 226-229, 2007.
[40] Cheng, Y. Mean-shift, Mode Seeking and Clustering[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 17, pp. 790, 1995.
[41] Comaniciu D, Ramesh V, Meer P. Kernel-based Object Tracking[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 25, pp. 564-577, 2003.
[42] Djouadi A,Snorrason 0,Garber F D. The Quality of Training-Sample Estimates of the Bhattacharyya Coefficient[J].IEEE Trans on Pattern Analysis and Machine Intelligent,1990,12:92—97.
[43] Nummiaro K, Koller-Meier E and Gool L.V. An Adaptive Color-based Particle Filter[J]. Image and Vision Computing, Vol. 21, pp. 99–110, 2003.
[44] Yang C, Duraiswami R and Davis L. Fast Multiple Object Tracking via a Hierarchical Particle Filter[J]. In The Proceedings of the Tenth IEEE International Conference on Computer Vision, (ICCV 2005), Beijing, China, pp. 212–219, 2005.
[45] Joachims Y. Estimating the Generalization Performance of a SVM Efficiently[A], in Proceedings of the International Conference on Machine Lerning[C], Morgan Kaufman,2000