|
|
- 2018
基于鉴别字典学习的遮挡人脸姿态识别
|
Abstract:
利用字典学习与稀疏表示的信号重建与分类的性能，两步字典训练学习方法引入到鲁棒性人脸姿态识别中。首先，将人脸姿态离散化为不同的子空间，使用K-奇异值分解法（K-SVD）为每个子空间训练一个子字典使其对应一个类别；然后，将所有子字典组合成超完备字典；最后，采用基于Gabor特征与稀疏表示的方法进行姿态分类。为了提高字典的分类能力，本文采用两步字典训练学习方法，并在第二步学习中加入类别约束；为了提高算法的鲁棒性，本文重构一个遮挡人脸字典，解决人脸姿态识别中人脸遮挡问题。通过在公开的XJTU、PIE和CAS-PEAL-R1人脸库上的实验结果表明，本文方法在具有光照、噪声和遮挡变化的人脸库识别率均能达到95%左右，基本能达到实际应用的要求
| [1] | The CMU PIE database[EB/OL]. http://www.ri.cmu.edu/projects/project_418.html, 2013 |
| [2] | Gao W, Cao B, Shan S G,et al. The CAS-PEAL Large-Scale Chinese Face Database and Baseline Evaluations[J]. <em>IEEE Transaction on System Man, and Cybernetics (Part A), </em>2008, 38:149-161 |
| [3] | Dou P F, Wu Y H, Shah S K, et al. Benchmarking 3D Pose Estimation for Face Recognition[C]. International Conference on Pattern Recognition, Stockholm, Sweden, 2014 |
| [4] | Akos P, Levente H. Precise 3D Pose Estimation of Human Faces[C]. The 9th International Conference on Computer Vision Theory and Applications, Lisbon, Portugal, 2014 |
| [5] | Li D Q, Pedrycz W. A Central Profile-Based 3D Face Pose Estimation[J]. <em>Pattern Recognition</em>, 2014, 47(2):525-534 |
| [6] | Yang M, Zhang L, Shiu S C K, et al. Gabor Feature Based Robust Representation and Classification for Face Recognition with Gabor Occlusion Dictionary[J]. <em>Pattern Recognition</em>, 2013, 46(2):1865-1878 |
| [7] | Xiang Jinhai, Fan Heng, Xu Jun, et al. Object Tracking Based on Local Sparse Representation[J]. <em>J. Huazhong Univ. of Sci. & Tech. (Natural Science Edition), </em>2014, 42(7):187-193(向金海,樊恒,徐俊,等. 基于局部稀疏表示的目标跟踪[J]. 华中科技大学学报(自然科学版),2014, 42(7):187-193) |
| [8] | Wang C, Song X B. Robust Head Pose Estimation Using Supervised Manifold Projection[C]. International Conference on Image Processing, USA, 2012:161-164 |
| [9] | Mu？oz-Salinas R, Yeguas-Bolivar E, Saffiotti A, et al. Multi-camera Head Pose Estimation[J]. <em>Machine Vision and Applications</em>, 2012, 23(3):479-490 |
| [10] | Nuevo J, Bergasa L M, Jiménez P. RSMAT:Robust Simultaneous Modeling and Tracking[J].<em>Pattern Recognition Letters</em>, 2010, 31(16):2455-2463 |
| [11] | Chen Zhenxue, Chang Faliang, Liu Chunsheng, et al. Pose Parameters Estimate Based on AdaBoost Algorithm and Facial Feature Triangle[J]. <em>Geomatics and Information Science of Wuhan University</em>, 2011, 36(10):1164-1167(陈振学,常发亮,刘春生, 等. 基于Adaboost算法和人脸特征三角形的姿态参数估计[J].武汉大学学报·信息科学版,2011, 36(10):1164-1167) |
| [12] | Ma B P, Li A A, Chai X J, et al. CovGa:A novel Descriptor Based on Symmetry of Regions for Head Pose Estimation[J].<em>Neurocomputing</em>, 2014, 143:97-108 |
| [13] | Chen X P, Yang Q Q, Liao H H,et al. Real-Time Face Pose Estimation in Video Sequence[C]. The Second International Workshop on Education Technology and Computer Science, Wuhan, 2010 |
| [14] | Srinivasan S, Boyer K L. Head Pose Estimation Using View Based Eigenspaces[C]. The 16th International Conference on Pattern Recognition, Quebec City, Canada, 2002 |
| [15] | Lis Z. Learning Multi-view Face Subspaces and Facial Pose Estimation Using Independent Component Analysis[J].<em>IEEE Trans. Image Process,</em> 2005, 14(6):705-712 |
| [16] | Wu J, Trivedi M M. A Two-Stage Head Pose Estimation Framework and Valuation[J].<em>Pattern Recognition</em>, 2008, 41(5):1138-1158 |
| [17] | Zhang Y Y, Idrissi K, Garcia C. A Dictionary-learning Sparse Representation Framework for Pose Classification[C]. IEEE International Workshop on Machine Learning for Signal Processing, UK, 2013 |
| [18] | Zhang Q, Li B. Discriminative K-SVD for Dictionary Learning in Face Recognition[C]. IEEE Conference on Computer Vision & Pattern Recognition, USA, 2010 |
| [19] | Meng Yang, Luc Van Gool, Lei Zhang. Sparse Variation Dictionary Learning for Face Recognition with a Single Sample Per Person[C]. IEEE International Conference on Computer Vision, San Francisco, USA, 2013 |
| [20] | AI & R Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University. Oriental Face Database[EB/OL]. http://www.aiar.xjtu.edu.cn/groups/face/Chinese/Homepage.htm, 2013 |
| [21] | Wang Huaqing. Face Pose Estimation with Local Geometry Preserving[D]. Xi'an:Xidian University, 2013(王华青.基于局部几何结构的人脸图像姿态估计[D].西安:西安电子科技大学,2013) |
