Searching for Variable-Speed Motions in Long Sequences of Motion Capture Data

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14330%2F19%3A00107162" target="_blank" >RIV/00216224:14330/19:00107162 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1016/j.is.2018.04.002" target="_blank" >http://dx.doi.org/10.1016/j.is.2018.04.002</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.is.2018.04.002" target="_blank" >10.1016/j.is.2018.04.002</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Searching for Variable-Speed Motions in Long Sequences of Motion Capture Data

Popis výsledku v původním jazyce

Motion capture data digitally represent human movements by sequences of body configurations in time. Subsequence searching in long sequences of such spatio-temporal data is difficult as query-relevant motions can vary in execution speeds and styles and can occur anywhere in a very long data sequence. To deal with these problems, we employ a fast and effective similarity measure that is elastic. The property of elasticity enables matching of two overlapping but slightly misaligned subsequences with a high confidence. Based on the elasticity, the long data sequence is partitioned into overlapping segments that are organized in multiple levels. The number of levels and sizes of overlaps are optimized to generate a modest number of segments while being able to trace an arbitrary query. In a retrieval phase, a query is always represented as a single segment and fast matched against segments within a relevant level without any costly post-processing. Moreover, visiting adjacent levels makes possible subsequence searching of time-warped (i.e., faster or slower executed) queries. To efficiently search on a large scale, segment features can be binarized and segmentation levels independently indexed. We experimentally demonstrate effectiveness and efficiency of the proposed approach for subsequence searching on a real-life dataset.

Název v anglickém jazyce

Searching for Variable-Speed Motions in Long Sequences of Motion Capture Data

Popis výsledku anglicky

Motion capture data digitally represent human movements by sequences of body configurations in time. Subsequence searching in long sequences of such spatio-temporal data is difficult as query-relevant motions can vary in execution speeds and styles and can occur anywhere in a very long data sequence. To deal with these problems, we employ a fast and effective similarity measure that is elastic. The property of elasticity enables matching of two overlapping but slightly misaligned subsequences with a high confidence. Based on the elasticity, the long data sequence is partitioned into overlapping segments that are organized in multiple levels. The number of levels and sizes of overlaps are optimized to generate a modest number of segments while being able to trace an arbitrary query. In a retrieval phase, a query is always represented as a single segment and fast matched against segments within a relevant level without any costly post-processing. Moreover, visiting adjacent levels makes possible subsequence searching of time-warped (i.e., faster or slower executed) queries. To efficiently search on a large scale, segment features can be binarized and segmentation levels independently indexed. We experimentally demonstrate effectiveness and efficiency of the proposed approach for subsequence searching on a real-life dataset.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10200 - Computer and information sciences

Projekt

<a href="/cs/project/GA16-18889S" target="_blank" >GA16-18889S: Analytika pro velká nestrukturovaná data</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2019

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Information Systems

ISSN

0306-4379

e-ISSN

—

Svazek periodika

80

Číslo periodika v rámci svazku

February

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

11

Strana od-do

148-158

Kód UT WoS článku

000454964800012

EID výsledku v databázi Scopus

2-s2.0-85045710372