We present a cascaded real-time system that recognizes dance patterns from 3D motion capture data. In a first step, the body trajectory, relative to the motion capture sensor, is matched. In a second step, an angular representation of the skeleton is proposed to make the system invariant to anthropometric differences relative to the body trajectory. Coping with non-uniform speed variations and amplitude discrepancies between dance patterns is achieved via a sequence similarity measure based on Dynamic Time Warping (DTW). A similarity threshold for recognition is automatically determined. Using only one good motion exemplar (baseline) per dance pattern, the recognition system is able to find a matching candidate pattern in a continuous stream of data, without prior segmentation. Experiments show the proposed algorithm reaches a good tradeoff between simplicity, speed and recognition rate. An average recognition rate of 86.8% is obtained in real-time.
Oveneke, M, Enescu, V & Sahli, H 2012, Real-time Dance Pattern Recognition Invariant to Anthropometric and Temporal Differences. in J Blanc-talon, W Philips, D Popescu, P Scheunders & P Zemcãk (eds), Lecture Notes in Computer Science. vol. 7517, Advanced Concepts for Intelligent Vision Systems, Springer Verlag, pp. 407-419, the 14th international conference on Advanced Concepts for Intelligent Vision Systems, Brno, Czech Republic, 4/09/12. <http://www.springerlink.com/content/p53178712240tl21/>
Oveneke, M., Enescu, V., & Sahli, H. (2012). Real-time Dance Pattern Recognition Invariant to Anthropometric and Temporal Differences. In J. Blanc-talon, W. Philips, D. Popescu, P. Scheunders, & P. Zemcãk (Eds.), Lecture Notes in Computer Science (Vol. 7517, pp. 407-419). (Advanced Concepts for Intelligent Vision Systems). Springer Verlag. http://www.springerlink.com/content/p53178712240tl21/
@inproceedings{20505fe3f46c494a9f20d517e69d3ee6,
title = "Real-time Dance Pattern Recognition Invariant to Anthropometric and Temporal Differences",
abstract = "We present a cascaded real-time system that recognizes dance patterns from 3D motion capture data. In a first step, the body trajectory, relative to the motion capture sensor, is matched. In a second step, an angular representation of the skeleton is proposed to make the system invariant to anthropometric differences relative to the body trajectory. Coping with non-uniform speed variations and amplitude discrepancies between dance patterns is achieved via a sequence similarity measure based on Dynamic Time Warping (DTW). A similarity threshold for recognition is automatically determined. Using only one good motion exemplar (baseline) per dance pattern, the recognition system is able to find a matching candidate pattern in a continuous stream of data, without prior segmentation. Experiments show the proposed algorithm reaches a good tradeoff between simplicity, speed and recognition rate. An average recognition rate of 86.8% is obtained in real-time.",
keywords = "dance motion recognition, real-time, similarity measure, threshold determination, Dynamic Time Warping (DTW)",
author = "Meshia Oveneke and Valentin Enescu and Hichem Sahli",
note = "Jacques Blanc-Talon, Wilfried Philips, Dan Popescu, Paul Scheunders, Pavel Zemc{\'i}k; the 14th international conference on Advanced Concepts for Intelligent Vision Systems, ACIVS 2012 ; Conference date: 04-09-2012 Through 07-09-2012",
year = "2012",
month = sep,
day = "4",
language = "English",
isbn = "978-3-642-33139-8",
volume = "7517",
series = "Advanced Concepts for Intelligent Vision Systems",
publisher = "Springer Verlag",
pages = "407--419",
editor = "Jacques Blanc-talon and Wilfried Philips and Dan Popescu and Paul Scheunders and Pavel Zemc{\~a}k",
booktitle = "Lecture Notes in Computer Science",
address = "Germany",
}