Deterministic 3D human pose estimation using rigid structure

作者: Jack Valmadre , Simon Lucey

DOI: 10.1007/978-3-642-15558-1_34

关键词:

摘要: This paper explores a method, first proposed by Wei and Chai [1], for estimating 3D human pose from several frames of uncalibrated 2D point correspondences containing projected body joint locations. In their work boldly claimed that, through the introduction rigid constraints to torso hip, camera scales, bone lengths absolute depths could be estimated finite number (i.e. ≥ 5). this we show claim false, demonstrating in principle one can never estimate these parameters frames. Further, demonstrate approach is only valid sub-structures (e.g. torso). Based on analysis propose novel using deterministic structure motion based assumptions rigidity body's torso. Our provides notably more accurate estimates substantially faster than Chai's approach, unlike original, solved as least-squares problem.

参考文章(14)
I. van Vajk, D. Chetverikov, L. Hajder, Robust structure from motion under weak perspective international symposium on 3d data processing visualization and transmission. pp. 828- 835 ,(2004) , 10.1109/3DPVT.2004.110
Jing Xiao, Jinxiang Chai, Takeo Kanade, A Closed-Form Solution to Non-Rigid Shape and Motion Recovery International Journal of Computer Vision. ,vol. 67, pp. 233- 246 ,(2006) , 10.1007/S11263-005-3962-9
C. Barron, I.A. Kakadiaris, Estimating anthropometry and pose from a single image computer vision and pattern recognition. ,vol. 1, pp. 669- 676 ,(2000) , 10.1109/CVPR.2000.855884
L. Torresani, A. Hertzmann, C. Bregler, Nonrigid Structure-from-Motion: Estimating Shape and Motion with Hierarchical Priors IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 30, pp. 878- 892 ,(2008) , 10.1109/TPAMI.2007.70752
C. Bregler, A. Hertzmann, H. Biermann, Recovering non-rigid 3D shape from image streams computer vision and pattern recognition. ,vol. 2, pp. 690- 696 ,(2000) , 10.1109/CVPR.2000.854941
T. Morita, T. Kanade, A sequential factorization method for recovering shape and motion from image streams IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 19, pp. 858- 867 ,(1997) , 10.1109/34.608289
Christoph Bregler, Lorenzo Torresani, Aaron Hertzmann, Learning Non-Rigid 3D Shape from 2D Motion neural information processing systems. ,vol. 16, pp. 1555- 1562 ,(2003)
C.J. Taylor, Reconstruction of articulated objects from point correspondences in a single uncalibrated image computer vision and pattern recognition. ,vol. 1, pp. 677- 684 ,(2000) , 10.1109/CVPR.2000.855885
Ijaz Akhter, Takeo Kanade, Yaser Sheikh, Sohaib Khan, Nonrigid Structure from Motion in Trajectory Space neural information processing systems. ,vol. 21, pp. 41- 48 ,(2008)
Carlo Tomasi, Takeo Kanade, Shape and motion from image streams under orthography: a factorization method International Journal of Computer Vision. ,vol. 9, pp. 137- 154 ,(1992) , 10.1007/BF00129684