计算机科学

计算机科学 ›› 2017, Vol. 44 ›› Issue (2): 306-308.doi: 10.11896/j.issn.1002-137X.2017.02.052

• 图形图像与模式识别 • 上一篇    下一篇

基于L2,1范数稀疏特征选择和超法向量的深度图像序列行为识别

宋相法,张延锋,郑逢斌   

  1. 河南大学计算机与信息工程学院 开封475004,河南大学计算机与信息工程学院 开封475004,河南大学计算机与信息工程学院 开封475004
  • 出版日期:2018-11-13 发布日期:2018-11-13
  • 基金资助:
    本文受国家自然科学基金(U1504611,61272282),河南省教育厅科学技术研究重点项目(15A520010)资助

Activity Recognition from Depth Image Sequences Based on L2,1-norm Sparse Feature Selection and Super Normal Vector

SONG Xiang-fa, ZHANG Yan-feng and ZHENG Feng-bin   

  • Online:2018-11-13 Published:2018-11-13

PDF (PC)

617

摘要： 结合L2,1范数稀疏特征选择和超法向量提出了一种新的深度图像序列行为识别方法。首先从深度图像序列中提取超法向量特征；然后利用L2,1范数稀疏特征选择方法从超法向量特征中选择出最具判别性的稀疏特征子集作为特征表示；最后利用线性分类器Liblinear进行分类。在MSR Action3D数据库上的实验结果表明,所提方法使用2%的超法向量特征获得的识别率为94.55%,并且 具有比 其他方法更高的识别精度。

关键词: 行为识别,深度图像序列,超法向量,稀疏特征选择,L2,1范数

Abstract: This paper presented a novel method of activity recognition from depth image sequences based on L2,1-norm sparse feature selection and super normal vector.First,the super normal vector feature is extracted from depth image sequences.Then the most discriminative feature subset is selected from the whole super normal vector feature set based on the method of L2,1-norm sparse feature selection.Finally,the classification is based on Liblinear classifier.Experimental results on MSR Action3D dataset show that the proposed method achieves 94.55% of recognition accuracy using only 2% of the whole super normal vector feature,and is superior to the state-of-art methods.

Key words: Activity recognition,Depth image sequences,Super normal vector,Sparse feature selection,L2,1-norm

[1] AGGARWAL J K,RYOO M S.Human activity analysis:a review[J].ACM Computing Surveys,2011,3(3):1-43.
[2] HU Q,QIN L,HUANG Q M.A survey on visual human action recognition[J].Chinese Journal of Computers,2013,6(12):2512-2524.(in Chinese) 胡琼,秦磊,黄庆明.基于视觉的人体动作识别综述[J].计算机学报,2013,6(12):2512-2524.
[3] HAN J H,SHAO L,XU D,et al.Enhanced computer visionwith microsoft kinect sensor:a review[J].IEEE Transactions on Cybernetics,2013,43(5):1318-1334.
[4] AGGARWAL J K,LU X.Human activity recognition from 3D data:a review[J].Pattern Recognition Letters,2014,8(2):70-80.
[5] LEI Q,CHEN D S,LI S Z.Advances on human action recognition in realistic scenes[J].Computer Science,2014,41(12):1-7.(in Chinese) 雷庆,陈锻生,李绍滋.复杂场景下的人体行为识别研究新进展[J].计算机科学,2014,41(12):1-7.
[6] LI W Q,ZHANG Z Y,LIU Z C.Action recognition based on a bag of 3D points[C]∥Proceedings of the IEEE International Conference on Human Communicative Behavior Analysis.2010:9-14.
[7] WANG J,LIU Z C,WU Y,et al.Mining actionlet ensemble for action recognition with depth cameras[C]∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.2012:1291-1297.
[8] YANG X D,ZHANG C Y,TIAN Y L.Recognition actions using depth motion maps-based histograms of oriented gradients[C]∥Proceedings of ACM Conference on Multimedia.2012:1057-1060.
[9] LU X,AGGARWAL J K.Spatio-temporal depth cuboid simila-rity feature for activity recognition using depth camera[C]∥Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition.2013:2834-2841.
[10] OREIFEJ O,LIU Zi C.HON4D:Histogram of oriented 4D normals for activity recognition from depth sequences[C]∥Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition.2013:716-723.
[11] YANG X D,TIAN Y L.Super normal vector for activity recognition using depth sequences[C]∥Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition.2014:804-811.
[12] SHEN X X,ZHANG H,GAO Z,et al.Behavior recognition algorithm based on depth information and RGB Image[J].Pattern Recognition and Artificial Intelligence,2013,26(8):722-728.(in Chinese) 申晓霞,张桦,高赞,等.基于深度信息和RGB图像的行为识别算法[J].模式识别与人工智能,2013,6(8):722-728.
[13] WANG X,WO B H,GUAN Q,et al.Human action recognition based on manifold learning[J].Journal of Image and Graphics,2014,19(6):914-923.(in Chinese) 王鑫,沃波海,管秋,等.基于流形学习的人体动作识别[J].中国图象图形学报,2014,9(6):914-923.
[14] NIE F P,HUANG H,CAI X,et al.Efficient and robust feature selection via joint L2,1-norms minimization[C]∥Proceedings of International Conference on Neural Information Processing Systems.2010:1813-1821.
[15] JORGE S,FLORENT P,THOMAS M,et al.Image classification with the fisher vector:theory and practice[J].International Journal of Computer Vision,2013,5(3):222-245.
[16] MAIRAL J,BACH F,PONCE J,et al.Online learning for matrix factorization and sparse coding[J].Journal of Machine Learning Research,2010,11(1):19-60.
[17] HE R,TAN T N,WANG L,et al.L2,1 Regularized correntropy for robust feature selection[C]∥Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition.2012:2504-2511.
[18] SHI X S,YANG Y J,GUO Z H,et al.Face recognition bysparse discriminant analysis via joint L2,1-norm minimization[J].Pattern Recognition,2014,7(7):2447-2453.
[19] SHI C J,RUAN Q Q.Feature selection with enhanced sparsity for web image annotation[J].Journal of Software,2015,26(7):1800-1811.(in Chinese) 史彩娟,阮秋琦.基于增强稀疏性特征选择的网络图像标注[J].软件学报,2015,6(7):1800-1811.
[20] ZHOU P Y,LI J,SHEN N M,et al.BSFCoS:Fast co-saliency detection based on block and sparse principal feature extraction[J].Computer Science,2015,2(8):305-309.(in Chinese) 周培云,李静,沈宁敏,等.BSFCoS:基于分块与稀疏主特征提取的快速协同显著性检测[J].计算机科学,2015,2(8):305-309.
[21] FAN R G,CHANG K W,HSIEH C J,et al.LIBLINEAR:A library for large linear classification[J].Journal of Machine Learning Research,2008,9(8):1871-1874.
No related articles found!
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发