{"title":"Relevant LMA Features for Human Motion Recognition","authors":"Insaf Ajili, Malik Mallem, Jean-Yves Didier","volume":141,"journal":"International Journal of Computer and Information Engineering","pagesStart":792,"pagesEnd":797,"ISSN":"1307-6892","URL":"https:\/\/publications.waset.org\/pdf\/10009574","abstract":"Motion recognition from videos is actually a very
\r\ncomplex task due to the high variability of motions. This paper
\r\ndescribes the challenges of human motion recognition, especially
\r\nmotion representation step with relevant features. Our descriptor
\r\nvector is inspired from Laban Movement Analysis method. We
\r\npropose discriminative features using the Random Forest algorithm
\r\nin order to remove redundant features and make learning algorithms
\r\noperate faster and more effectively. We validate our method on
\r\nMSRC-12 and UTKinect datasets.","references":"1] I. Ajili, M. Mallem, and J. Y. Didier. Gesture recognition for humanoid\r\nrobot teleoperation. In 2017 26th IEEE International Symposium\r\non Robot and Human Interactive Communication (RO-MAN), pages\r\n1115\u20131120, Aug 2017.\r\n[2] I. Ajili, M. Mallem, and J.-Y. Didier. Robust human action\r\nrecognition system using laban movement analysis. Procedia Computer\r\nScience, 112(Supplement C):554 \u2013 563, 2017. Knowledge-Based and\r\nIntelligent Information & Engineering Systems: Proceedings of the\r\n21st International Conference, KES-20176-8 September 2017, Marseille,\r\nFrance.\r\n[3] C. B. Barber, D. P. Dobkin, and H. Huhdanpaa. The quickhull algorithm\r\nfor convex hulls. ACM Trans. Math. Softw., 22(4):469\u2013483, Dec. 1996.\r\n[4] L. Breiman. Random forests. Mach. Learn., 45(1):5\u201332, Oct. 2001.\r\n[5] A. B.Surendiran1. Feature selection using stepwise anova discriminant\r\nanalysis for mammogram mass classification. International Journal on\r\nSignal & Image Processing, 2(1):4, January 2011.\r\n[6] S. Fothergill, H. Mentis, P. Kohli, and S. Nowozin. Instructing people\r\nfor training gestural interactive systems. In Proceedings of the SIGCHI\r\nConference on Human Factors in Computing Systems, CHI \u201912, pages\r\n1737\u20131746, New York, NY, USA, 2012. ACM.\r\n[7] M. E. Hussein, M. Torki, M. A. Gowayyed, and M. El-Saban. Human\r\naction recognition using a temporal hierarchy of covariance descriptors\r\non 3d joint locations. In Proceedings of the Twenty-Third International\r\nJoint Conference on Artificial Intelligence, IJCAI \u201913, pages 2466\u20132472.\r\nAAAI Press, 2013.\r\n[8] I. Laptev and T. Lindeberg. Space-time interest points. In Computer\r\nVision, 2003. Proceedings. Ninth IEEE International Conference on,\r\npages 432\u2013439. IEEE, 2003.\r\n[9] C. Lazar, J. Taminau, S. Meganck, D. Steenhoff, A. Coletta, C. Molter,\r\nV. de Schaetzen, R. Duque, H. Bersini, and A. Nowe. A survey on filter\r\ntechniques for feature selection in gene expression microarray analysis.\r\nIEEE\/ACM Trans. Comput. Biol. Bioinformatics, 9(4):1106\u20131119, July\r\n2012.\r\n[10] A. M. Lehrmann, P. V. Gehler, and S. Nowozin. Efficient nonlinear\r\nmarkov models for human motion. In 2014 IEEE Conference on\r\nComputer Vision and Pattern Recognition, pages 1314\u20131321, June 2014.\r\n[11] H. Wang, A. Kl\u00a8aser, C. Schmid, and C.-L. Liu. Dense trajectories and\r\nmotion boundary descriptors for action recognition. Int. J. Comput. Vis.,\r\n103(1):60\u201379, 2013.\r\n[12] P. Wang, Z. Li, Y. Hou, and W. Li. Action recognition based on\r\njoint trajectory maps using convolutional neural networks. CoRR,\r\nabs\/1611.02447, 2016.\r\n[13] J. Weston, S. Mukherjee, O. Chapelle, M. Pontil, T. Poggio, and\r\nV. Vapnik. Feature selection for svms. In Proceedings of the 13th\r\nInternational Conference on Neural Information Processing Systems,\r\nNIPS\u201900, pages 647\u2013653, Cambridge, MA, USA, 2000. MIT Press.\r\n[14] L. Xia, C. C. Chen, and J. K. Aggarwal. View invariant human action\r\nrecognition using histograms of 3d joints. In 2012 IEEE Computer\r\nSociety Conference on Computer Vision and Pattern Recognition\r\nWorkshops, pages 20\u201327, June 2012.\r\n[15] M. Yamada, W. Jitkrittum, L. Sigal, E. P. Xing, and M. Sugiyama.\r\nHigh-dimensional feature selection by feature-wise non-linear lasso.\r\nArXiv e-prints, Feb. 2012.\r\n[16] L. Zhou, W. Li, Y. Zhang, P. Ogunbona, D. T. Nguyen, and H. Zhang.\r\nDiscriminative key pose extraction using extended lc-ksvd for action\r\nrecognition. In 2014 International Conference on Digital Image\r\nComputing: Techniques and Applications (DICTA), pages 1\u20138, Nov\r\n2014.","publisher":"World Academy of Science, Engineering and Technology","index":"Open Science Index 141, 2018"}