Scalable and Compact Representation for Motion Capture Data Using Tensor Decomposition

Motion capture (mocap) technology is widely used in movie and game industries. Compact representation of the mocap data is critical to efficient storage and transmission. In this letter, we propose a novel tensor decomposition based scheme for compact and progressive representation of the mocap data. Our method segments and stacks the mocap sequence locally, and generates a 3rd-order tensor, which has strong correlation within and across slices of the tensor. Then, our method iteratively applies tensor decomposition in a multi-layer structure to explore the correlation characteristic. Experimental results demonstrate that the proposed scheme significantly outperforms existing algorithms in terms of scalability and storage requirement.

[1]  Tamara G. Kolda,et al.  Tensor Decompositions and Applications , 2009, SIAM Rev..

[2]  Lap-Pui Chau,et al.  A Fuzzy Clustering Algorithm for Virtual Character Animation Representation , 2011, IEEE Transactions on Multimedia.

[3]  Ivan V. Bajic,et al.  MoCap data coding with unrestricted quantization and rate control , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[4]  Kang Li,et al.  Model-Based Power Aware Compression Algorithms for MPEG-4 Virtual Human Animation in Mobile Environments , 2007, IEEE Transactions on Multimedia.

[5]  Andrzej Cichocki,et al.  Canonical Polyadic Decomposition Based on a Single Mode Blind Source Separation , 2012, IEEE Signal Processing Letters.

[6]  Lap-Pui Chau,et al.  Keyframe selection for motion capture using motion activity analysis , 2012, 2012 IEEE International Symposium on Circuits and Systems.

[7]  Yongwon Jeong.  Acoustic Model Adaptation Based on Tensor Analysis of Training Models , 2011, IEEE Signal Processing Letters.

[8]  Rynson W. H. Lau,et al.  The alpha parallelogram predictor: A lossless compression method for motion capture data , 2013, Inf. Sci..

[9]  Yi Li,et al.  Learning shift-invariant sparse representation of actions , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[10]  Erik G. Larsson,et al.  The Higher-Order Singular Value Decomposition: Theory and an Application [Lecture Notes] , 2010, IEEE Signal Processing Magazine.

[11]  Irene Cheng,et al.  Perceptually Guided Fast Compression of 3-D Motion Capture Data , 2011, IEEE Transactions on Multimedia.

[12]  Dan Schonfeld,et al.  Robust HOSVD-Based Higher-Order Data Indexing and Retrieval , 2013, IEEE Signal Processing Letters.

[13]  Daniel M. Dunlavy,et al.  A scalable optimization approach for fitting canonical tensor decompositions , 2011 .

[14]  Craig Gotsman,et al.  Compression of soft-body animation sequences , 2004, Comput. Graph..

[15]  Mario Fernando Montenegro Campos,et al.  Online gesture recognition from pose kernel learning and decision forests , 2014, Pattern Recognit. Lett..

[16]  Philippe Beaudoin,et al.  Adapting wavelet compression to human motion capture clips , 2007, GI '07.

[17]  Zhigang Deng,et al.  Compression of Human Motion Capture Data Using Motion Pattern Indexing , 2009, Comput. Graph. Forum.

[18]  Junhui Hou,et al.  Human motion capture data recovery using trajectory-based matrix completion , 2013 .