A Quick Filtering for Similarity Queries in Motion Capture Databases

A similarity retrieval of motion capture data has received substantial attention in recent years. In this paper, we focus on feature extraction and quick filtering methods in the similarity retrieval system. A representation of motion capture data is joint angles, which can distinguish different human body poses. We propose a new technique for dimensionality reduction based the average and variance of joint angles. Our dimensionality reduction is simple to understand and implement. In experiments, twenty dance motion clips each of which is different in length and style, are used in the test data set with a total of 60,000 frames. The results of our quick filtering show an achievement on the recall and precision up to 100% and 70%, respectively.

[1]  Eamonn J. Keogh,et al.  Scaling up dynamic time warping for datamining applications , 2000, KDD '00.

[2]  Chih-Yi Chiu,et al.  Content-based retrieval for human motion data , 2004, J. Vis. Commun. Image Represent..

[3]  David G. Stork,et al.  Pattern Classification , 1973 .

[4]  David Burshtein,et al.  An enhanced dynamic time warping model for improved estimation of DTW parameters , 2003, IEEE Trans. Speech Audio Process..

[5]  Osamu Hasegawa,et al.  Motion recognition based on Dynamic-Time Warping method with Self-Organizing Incremental Neural Network , 2008, 2008 19th International Conference on Pattern Recognition.

[6]  Kiyoko Yokoyama,et al.  Analysis of Japanese dance movements using motion capture system , 2006 .

[7]  Kozaburo Hachimura,et al.  Segmentation of Dancing Movement by Extracting Features from Motion Capture Data , 2008 .

[8]  Tido Röder,et al.  Efficient content-based retrieval of motion capture data , 2005, SIGGRAPH 2005.

[9]  Eamonn J. Keogh,et al.  A Simple Dimensionality Reduction Technique for Fast Similarity Search in Large Time Series Databases , 2000, PAKDD.

[10]  Dimitrios Gunopulos,et al.  Indexing Multidimensional Time-Series , 2004, The VLDB Journal.

[11]  Jernej Barbic,et al.  Segmenting Motion Capture Data into Distinct Behaviors , 2004, Graphics Interface.

[12]  Kozaburo Hachimura,et al.  Comparison of Structural Variables with Spatio-temporal Variables Concerning the Identifiability of Okuri Class and Player in Japanese Traditional Dancing , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[13]  Toru Nakata.  Temporal segmentation and recognition of body motion data based on inter-limb correlation analysis , 2007, 2007 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[14]  Mamiko Sakata,et al.  Quantification of dance movement by simultaneous measurement of body motion and biophysical information , 2007, Int. J. Autom. Comput..