Enriching a motion database by analogous combination of partial human motions

Won Seob Jang, Won Kyu Lee, In Kwon Lee, Jehee Lee

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

We have synthesized new human body motions from existing motion data, by dividing the body of an animated character into several parts, such as upper and lower body, and partitioning the motion of the character into corresponding partial motions. By combining different partial motions, we can generate new motion sequences. We select the most natural-looking combinations by analyzing the similarity of partial motions, using techniques such as motion segmentation, dimensionality reduction, and clustering. These new combinations can dramatically increase the size of a motion database, allowing more score in selecting motions to meet constraints, such as collision avoidance. We verify the naturalness and physical plausibility of the new motions using an SVM learning model and by analysis of static and dynamic balance.

Original languageEnglish
Pages (from-to)271-280
Number of pages10
JournalVisual Computer
Volume24
Issue number4
DOIs
Publication statusPublished - 2008 Apr 1

Fingerprint

Collision avoidance

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Vision and Pattern Recognition
  • Computer Graphics and Computer-Aided Design

Cite this

Jang, Won Seob ; Lee, Won Kyu ; Lee, In Kwon ; Lee, Jehee. / Enriching a motion database by analogous combination of partial human motions. In: Visual Computer. 2008 ; Vol. 24, No. 4. pp. 271-280.
@article{f880a37cd7814da1af45b4075482756f,
title = "Enriching a motion database by analogous combination of partial human motions",
abstract = "We have synthesized new human body motions from existing motion data, by dividing the body of an animated character into several parts, such as upper and lower body, and partitioning the motion of the character into corresponding partial motions. By combining different partial motions, we can generate new motion sequences. We select the most natural-looking combinations by analyzing the similarity of partial motions, using techniques such as motion segmentation, dimensionality reduction, and clustering. These new combinations can dramatically increase the size of a motion database, allowing more score in selecting motions to meet constraints, such as collision avoidance. We verify the naturalness and physical plausibility of the new motions using an SVM learning model and by analysis of static and dynamic balance.",
author = "Jang, {Won Seob} and Lee, {Won Kyu} and Lee, {In Kwon} and Jehee Lee",
year = "2008",
month = "4",
day = "1",
doi = "10.1007/s00371-007-0200-1",
language = "English",
volume = "24",
pages = "271--280",
journal = "Visual Computer",
issn = "0178-2789",
publisher = "Springer Verlag",
number = "4",

}

Enriching a motion database by analogous combination of partial human motions. / Jang, Won Seob; Lee, Won Kyu; Lee, In Kwon; Lee, Jehee.

In: Visual Computer, Vol. 24, No. 4, 01.04.2008, p. 271-280.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Enriching a motion database by analogous combination of partial human motions

AU - Jang, Won Seob

AU - Lee, Won Kyu

AU - Lee, In Kwon

AU - Lee, Jehee

PY - 2008/4/1

Y1 - 2008/4/1

N2 - We have synthesized new human body motions from existing motion data, by dividing the body of an animated character into several parts, such as upper and lower body, and partitioning the motion of the character into corresponding partial motions. By combining different partial motions, we can generate new motion sequences. We select the most natural-looking combinations by analyzing the similarity of partial motions, using techniques such as motion segmentation, dimensionality reduction, and clustering. These new combinations can dramatically increase the size of a motion database, allowing more score in selecting motions to meet constraints, such as collision avoidance. We verify the naturalness and physical plausibility of the new motions using an SVM learning model and by analysis of static and dynamic balance.

AB - We have synthesized new human body motions from existing motion data, by dividing the body of an animated character into several parts, such as upper and lower body, and partitioning the motion of the character into corresponding partial motions. By combining different partial motions, we can generate new motion sequences. We select the most natural-looking combinations by analyzing the similarity of partial motions, using techniques such as motion segmentation, dimensionality reduction, and clustering. These new combinations can dramatically increase the size of a motion database, allowing more score in selecting motions to meet constraints, such as collision avoidance. We verify the naturalness and physical plausibility of the new motions using an SVM learning model and by analysis of static and dynamic balance.

UR - http://www.scopus.com/inward/record.url?scp=41149088316&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=41149088316&partnerID=8YFLogxK

U2 - 10.1007/s00371-007-0200-1

DO - 10.1007/s00371-007-0200-1

M3 - Article

AN - SCOPUS:41149088316

VL - 24

SP - 271

EP - 280

JO - Visual Computer

JF - Visual Computer

SN - 0178-2789

IS - 4

ER -