Featureweighting in dynamic timewarping for gesture recognition in depth data

We present a gesture recognition approach for depth video data based on a novel Feature Weighting approach within the Dynamic Time Warping framework. Depth features from human joints are compared through video sequences using Dynamic Time Warping, and weights are assigned to features based on inter-intra class gesture variability. Feature Weighting in Dynamic Time Warping is then applied for recognizing begin-end of gestures in data sequences. The obtained results recognizing several gestures in depth data show high performance compared with classical Dynamic Time Warping approach.

[1]  Jessica K. Hodgins,et al.  Aligned Cluster Analysis for temporal segmentation of human motion , 2008, 2008 8th IEEE International Conference on Automatic Face & Gesture Recognition.

[2]  Tanja Schultz,et al.  HMM-based human motion recognition with optical flow data , 2009, 2009 9th IEEE-RAS International Conference on Humanoid Robots.

[3]  Honghai Liu,et al.  Recognition of Human Motion From Qualitative Normalised Templates , 2007, J. Intell. Robotic Syst..

[4]  Marc Parizeau,et al.  A Comparative Analysis of Regional Correlation, Dynamic Time Warping, and Skeletal Tree Matching for Signature Verification , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Sukhendu Das,et al.  Real-Time Upper-Body Human Pose Estimation Using a Depth Camera , 2011, MIRAGE.

[6]  Linda G. Shapiro,et al.  A SIFT descriptor with global context , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[7]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[8]  Sebastian Thrun,et al.  Real time motion capture using a single time-of-flight camera , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[9]  Dragomir Anguelov,et al.  Object Pose Detection in Range Scan Data , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[10]  Jake K. Aggarwal,et al.  Segmentation of 3D range images using pyramidal data structures , 1990, [1990] Proceedings Third International Conference on Computer Vision.

[11]  Andrew W. Fitzgibbon,et al.  Real-time human pose recognition in parts from single depth images , 2011, CVPR 2011.

[12]  Cristian Sminchisescu,et al.  Conditional models for contextual human motion recognition , 2006, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.