Fast Subsequence Matching in Motion Capture Data

D 2017

Fast Subsequence Matching in Motion Capture Data

SEDMIDUBSKÝ, Jan, Pavel ZEZULA and Jan ŠVEC

Basic information

Original name

Fast Subsequence Matching in Motion Capture Data

Authors

SEDMIDUBSKÝ, Jan (203 Czech Republic, guarantor, belonging to the institution), Pavel ZEZULA (203 Czech Republic, belonging to the institution) and Jan ŠVEC (203 Czech Republic)

Edition

Cham, 21st European Conference on Advances in Databases and Information Systems, p. 59-72, 14 pp. 2017

Publisher

Springer

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

printed version "print"

RIV identification code

RIV/00216224:14330/17:00094760

Organization unit

Faculty of Informatics

ISBN

978-3-319-66916-8

DOI

http://dx.doi.org/10.1007/978-3-319-66917-5_5

UT WoS

000463611400005

Keywords in English

subsequence matching; motion capture data; content-based retrieval; similarity measure; segmentation; indexing

Abstract

V originále

Motion capture data digitally represent human movements by sequences of body configurations in time. Subsequence matching in such spatio-temporal data is difficult as query-relevant motions can vary in lengths and occur arbitrarily in a very long motion. To deal with these problems, we propose a new subsequence matching approach which (1) partitions both short query and long data motion into fixed-size segments that overlap only partly, (2) uses an effective similarity measure to efficiently retrieve data segments that are the most similar to query segments, and (3) localizes the most query-relevant subsequences within extended and merged retrieved segments in a four-step postprocessing phase. The whole retrieval process is effective and fast in comparison with related work. A real-life 68-minute data motion can be searched in about 1s with the average precision of 87.98% for 5-NN queries.

Links

GBP103/12/G084, research and development project

Name: Centrum pro multi-modální interpretaci dat velkého rozsahu

Investor: Czech Science Foundation

Citovat

SEDMIDUBSKÝ, Jan, Pavel ZEZULA and Jan ŠVEC. Fast Subsequence Matching in Motion Capture Data. In 21st European Conference on Advances in Databases and Information Systems. Cham: Springer, 2017, p. 59-72. ISBN 978-3-319-66916-8. Available from: https://dx.doi.org/10.1007/978-3-319-66917-5_5.

@inproceedings{1381785,
   author = {Sedmidubský, Jan and Zezula, Pavel and Švec, Jan},
   address = {Cham},
   booktitle = {21st European Conference on Advances in Databases and Information Systems},
   doi = {http://dx.doi.org/10.1007/978-3-319-66917-5_5},
   keywords = {subsequence matching; motion capture data; content-based retrieval; similarity measure; segmentation; indexing},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Cham},
   isbn = {978-3-319-66916-8},
   pages = {59-72},
   publisher = {Springer},
   title = {Fast Subsequence Matching in Motion Capture Data},
   year = {2017}
}

TY  - JOUR
ID  - 1381785
AU  - Sedmidubský, Jan - Zezula, Pavel - Švec, Jan
PY  - 2017
TI  - Fast Subsequence Matching in Motion Capture Data
PB  - Springer
CY  - Cham
SN  - 9783319669168
KW  - subsequence matching
KW  - motion capture data
KW  - content-based retrieval
KW  - similarity measure
KW  - segmentation
KW  - indexing
N2  - Motion capture data digitally represent human movements by sequences of body configurations in time. Subsequence matching in such spatio-temporal data is difficult as query-relevant motions can vary in lengths and occur arbitrarily in a very long motion. To deal with these problems, we propose a new subsequence matching approach which (1) partitions both short query and long data motion into fixed-size segments that overlap only partly, (2) uses an effective similarity measure to efficiently retrieve data segments that are the most similar to query segments, and (3) localizes the most query-relevant subsequences within extended and merged retrieved segments in a four-step postprocessing phase. The whole retrieval process is effective and fast in comparison with related work. A real-life 68-minute data motion can be searched in about 1s with the average precision of 87.98% for 5-NN queries.
ER  -

SEDMIDUBSKÝ, Jan, Pavel ZEZULA and Jan ŠVEC. Fast Subsequence Matching in Motion Capture Data. In \textit{21st European Conference on Advances in Databases and Information Systems}. Cham: Springer, 2017, p.~59-72. ISBN~978-3-319-66916-8. Available from: https://dx.doi.org/10.1007/978-3-319-66917-5\_{}5.

Detailed Information on Publication Record