A Quick Filtering for Similarity Queries in Motion Capture Databases

AUTHORS: Worawat Choensawat, Woong Choi and Kozaburo Hachimura

ABSTRACT: 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.

Keywords: Similarity Retrieval, Motion Capture, Body Motion Data, Dimension Reduction, Minimal Bounding Envelop, Feature Extraction.

SOURCE: Choensawat, Worawat, Woong Choi, and Kozaburo Hachimura. "A quick filtering for similarity queries in motion capture databases." Pacific-Rim Conference on Multimedia. Springer Berlin Heidelberg, 2009.

LINK: https://link.springer.com/chapter/10.1007/978-3-642-10467-1_35

REFERENCES: 

MLA		Choensawat, Worawat, Woong Choi, and Kozaburo Hachimura. "A quick filtering for similarity queries in motion capture databases." Pacific-Rim Conference on Multimedia. Springer, Berlin, Heidelberg, 2009.
APA		Choensawat, W., Choi, W., & Hachimura, K. (2009, December). A quick filtering for similarity queries in motion capture databases. In Pacific-Rim Conference on Multimedia (pp. 404-415). Springer, Berlin, Heidelberg.
ISO 690		CHOENSAWAT, Worawat; CHOI, Woong; HACHIMURA, Kozaburo. A quick filtering for similarity queries in motion capture databases. In: Pacific-Rim Conference on Multimedia. Springer, Berlin, Heidelberg, 2009. p. 404-415.