Enriching a motion collection by transplanting limbs

Leslie Ikemoto, David Alexander Forsyth

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper describes a method that can significantly increase the size of a collection of motion observations by cutting limbs from one motion sequence and attaching them to another. Not all such transplants are successful, because correlations across the body are a significant feature of human motion. The method uses randomized search based around a set of rules to generate transplants that are (a) likely to be successful and (b) likely to enrich the existing motion collection. The resulting frames are annotated by a classifier to tell whether they look like human motion or not. We evaluate the method by obtaining motion demands from an application, synthesizing motions to meet those demands, and then scoring the synthesized motions. Motions synthesized using transplants are generally somewhat better than those synthesized without using transplants, because transplanting generates many frames quite close to the original frames, so that it is easier for the motion synthesis process to find a good path in the motion graph. Furthermore, we show classifier errors tend to have relatively little impact in practice. Finally, we show that transplanted motion data can be used to synthesize motions of a group coordinated in space and time without producing motions that share frames.

Original languageEnglish (US)
Title of host publicationComputer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation
PublisherAssociation for Computing Machinery, Inc
Pages99-108
Number of pages10
ISBN (Print)3905673142, 9783905673142
DOIs
StatePublished - Aug 27 2004
Externally publishedYes
Event2004 ACM SIGGRAPH / Eurographics Symposium on Computer Animation, SCA 2004 - Grenoble, France
Duration: Aug 27 2004Aug 29 2004

Publication series

NameComputer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation

Other

Other2004 ACM SIGGRAPH / Eurographics Symposium on Computer Animation, SCA 2004
CountryFrance
CityGrenoble
Period8/27/048/29/04

Fingerprint

Transplants
Classifiers

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Human-Computer Interaction
  • Software

Cite this

Ikemoto, L., & Forsyth, D. A. (2004). Enriching a motion collection by transplanting limbs. In Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation (pp. 99-108). (Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation). Association for Computing Machinery, Inc. https://doi.org/10.1145/1028523.1028537

Enriching a motion collection by transplanting limbs. / Ikemoto, Leslie; Forsyth, David Alexander.

Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation. Association for Computing Machinery, Inc, 2004. p. 99-108 (Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ikemoto, L & Forsyth, DA 2004, Enriching a motion collection by transplanting limbs. in Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation. Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation, Association for Computing Machinery, Inc, pp. 99-108, 2004 ACM SIGGRAPH / Eurographics Symposium on Computer Animation, SCA 2004, Grenoble, France, 8/27/04. https://doi.org/10.1145/1028523.1028537
Ikemoto L, Forsyth DA. Enriching a motion collection by transplanting limbs. In Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation. Association for Computing Machinery, Inc. 2004. p. 99-108. (Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation). https://doi.org/10.1145/1028523.1028537
Ikemoto, Leslie ; Forsyth, David Alexander. / Enriching a motion collection by transplanting limbs. Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation. Association for Computing Machinery, Inc, 2004. pp. 99-108 (Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation).
@inproceedings{313b387ea0bb4fc4a47f69662943003b,
title = "Enriching a motion collection by transplanting limbs",
abstract = "This paper describes a method that can significantly increase the size of a collection of motion observations by cutting limbs from one motion sequence and attaching them to another. Not all such transplants are successful, because correlations across the body are a significant feature of human motion. The method uses randomized search based around a set of rules to generate transplants that are (a) likely to be successful and (b) likely to enrich the existing motion collection. The resulting frames are annotated by a classifier to tell whether they look like human motion or not. We evaluate the method by obtaining motion demands from an application, synthesizing motions to meet those demands, and then scoring the synthesized motions. Motions synthesized using transplants are generally somewhat better than those synthesized without using transplants, because transplanting generates many frames quite close to the original frames, so that it is easier for the motion synthesis process to find a good path in the motion graph. Furthermore, we show classifier errors tend to have relatively little impact in practice. Finally, we show that transplanted motion data can be used to synthesize motions of a group coordinated in space and time without producing motions that share frames.",
author = "Leslie Ikemoto and Forsyth, {David Alexander}",
year = "2004",
month = "8",
day = "27",
doi = "10.1145/1028523.1028537",
language = "English (US)",
isbn = "3905673142",
series = "Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation",
publisher = "Association for Computing Machinery, Inc",
pages = "99--108",
booktitle = "Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation",

}

TY - GEN

T1 - Enriching a motion collection by transplanting limbs

AU - Ikemoto, Leslie

AU - Forsyth, David Alexander

PY - 2004/8/27

Y1 - 2004/8/27

N2 - This paper describes a method that can significantly increase the size of a collection of motion observations by cutting limbs from one motion sequence and attaching them to another. Not all such transplants are successful, because correlations across the body are a significant feature of human motion. The method uses randomized search based around a set of rules to generate transplants that are (a) likely to be successful and (b) likely to enrich the existing motion collection. The resulting frames are annotated by a classifier to tell whether they look like human motion or not. We evaluate the method by obtaining motion demands from an application, synthesizing motions to meet those demands, and then scoring the synthesized motions. Motions synthesized using transplants are generally somewhat better than those synthesized without using transplants, because transplanting generates many frames quite close to the original frames, so that it is easier for the motion synthesis process to find a good path in the motion graph. Furthermore, we show classifier errors tend to have relatively little impact in practice. Finally, we show that transplanted motion data can be used to synthesize motions of a group coordinated in space and time without producing motions that share frames.

AB - This paper describes a method that can significantly increase the size of a collection of motion observations by cutting limbs from one motion sequence and attaching them to another. Not all such transplants are successful, because correlations across the body are a significant feature of human motion. The method uses randomized search based around a set of rules to generate transplants that are (a) likely to be successful and (b) likely to enrich the existing motion collection. The resulting frames are annotated by a classifier to tell whether they look like human motion or not. We evaluate the method by obtaining motion demands from an application, synthesizing motions to meet those demands, and then scoring the synthesized motions. Motions synthesized using transplants are generally somewhat better than those synthesized without using transplants, because transplanting generates many frames quite close to the original frames, so that it is easier for the motion synthesis process to find a good path in the motion graph. Furthermore, we show classifier errors tend to have relatively little impact in practice. Finally, we show that transplanted motion data can be used to synthesize motions of a group coordinated in space and time without producing motions that share frames.

UR - http://www.scopus.com/inward/record.url?scp=84907370795&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84907370795&partnerID=8YFLogxK

U2 - 10.1145/1028523.1028537

DO - 10.1145/1028523.1028537

M3 - Conference contribution

AN - SCOPUS:84907370795

SN - 3905673142

SN - 9783905673142

T3 - Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation

SP - 99

EP - 108

BT - Computer Animation 2004 - ACM SIGGRAPH / Eurographics Symposium on Computer Animation

PB - Association for Computing Machinery, Inc

ER -