A data-driven method for real-time character animation in human-agent interaction

David Vogt, Steve Grehl, Erik Berger, Hani Ben Amor, Bernhard Jung

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

8 Citations (Scopus)

Abstract

We address the problem of creating believable animations for virtual humans that need to react to the body movements of a human interaction partner in real-time. Our data-driven approach uses prerecorded motion capture data of two interacting persons and performs motion adaptation during the live human-agent interaction. Extending the interaction mesh approach, our main contribution is a new scheme for efficient identification of motions in the prerecorded animation data that are similar to the live interaction. A global low-dimensional posture space serves to select the most similar interaction example, while local, more detail-rich posture spaces are used to identify poses closely matching the human motion. Using the interaction mesh of the selected motion example, an animation can then be synthesized that takes into account both spatial and temporal similarities between the prerecorded and live interactions.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages463-476
Number of pages14
Volume8637 LNAI
ISBN (Print)9783319097664
DOIs
StatePublished - 2014
Externally publishedYes
Event14th International Conference on Intelligent Virtual Agents, IVA 2014 - Boston, MA, United States
Duration: Aug 27 2014Aug 29 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8637 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other14th International Conference on Intelligent Virtual Agents, IVA 2014
CountryUnited States
CityBoston, MA
Period8/27/148/29/14

Fingerprint

Character Animation
Animation
Data-driven
Real-time
Interaction
Motion
Data acquisition
Mesh
Virtual Human
Motion Capture
Human
Person

Keywords

  • Character animation
  • interaction mesh
  • interactive characters
  • virtual agent

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Vogt, D., Grehl, S., Berger, E., Ben Amor, H., & Jung, B. (2014). A data-driven method for real-time character animation in human-agent interaction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8637 LNAI, pp. 463-476). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8637 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-09767-1_57

A data-driven method for real-time character animation in human-agent interaction. / Vogt, David; Grehl, Steve; Berger, Erik; Ben Amor, Hani; Jung, Bernhard.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8637 LNAI Springer Verlag, 2014. p. 463-476 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8637 LNAI).

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

Vogt, D, Grehl, S, Berger, E, Ben Amor, H & Jung, B 2014, A data-driven method for real-time character animation in human-agent interaction. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 8637 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8637 LNAI, Springer Verlag, pp. 463-476, 14th International Conference on Intelligent Virtual Agents, IVA 2014, Boston, MA, United States, 8/27/14. https://doi.org/10.1007/978-3-319-09767-1_57
Vogt D, Grehl S, Berger E, Ben Amor H, Jung B. A data-driven method for real-time character animation in human-agent interaction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8637 LNAI. Springer Verlag. 2014. p. 463-476. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-09767-1_57
Vogt, David ; Grehl, Steve ; Berger, Erik ; Ben Amor, Hani ; Jung, Bernhard. / A data-driven method for real-time character animation in human-agent interaction. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8637 LNAI Springer Verlag, 2014. pp. 463-476 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{d7da36ceffc24c0f94bb00c8800c36d4,
title = "A data-driven method for real-time character animation in human-agent interaction",
abstract = "We address the problem of creating believable animations for virtual humans that need to react to the body movements of a human interaction partner in real-time. Our data-driven approach uses prerecorded motion capture data of two interacting persons and performs motion adaptation during the live human-agent interaction. Extending the interaction mesh approach, our main contribution is a new scheme for efficient identification of motions in the prerecorded animation data that are similar to the live interaction. A global low-dimensional posture space serves to select the most similar interaction example, while local, more detail-rich posture spaces are used to identify poses closely matching the human motion. Using the interaction mesh of the selected motion example, an animation can then be synthesized that takes into account both spatial and temporal similarities between the prerecorded and live interactions.",
keywords = "Character animation, interaction mesh, interactive characters, virtual agent",
author = "David Vogt and Steve Grehl and Erik Berger and {Ben Amor}, Hani and Bernhard Jung",
year = "2014",
doi = "10.1007/978-3-319-09767-1_57",
language = "English (US)",
isbn = "9783319097664",
volume = "8637 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "463--476",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - A data-driven method for real-time character animation in human-agent interaction

AU - Vogt, David

AU - Grehl, Steve

AU - Berger, Erik

AU - Ben Amor, Hani

AU - Jung, Bernhard

PY - 2014

Y1 - 2014

N2 - We address the problem of creating believable animations for virtual humans that need to react to the body movements of a human interaction partner in real-time. Our data-driven approach uses prerecorded motion capture data of two interacting persons and performs motion adaptation during the live human-agent interaction. Extending the interaction mesh approach, our main contribution is a new scheme for efficient identification of motions in the prerecorded animation data that are similar to the live interaction. A global low-dimensional posture space serves to select the most similar interaction example, while local, more detail-rich posture spaces are used to identify poses closely matching the human motion. Using the interaction mesh of the selected motion example, an animation can then be synthesized that takes into account both spatial and temporal similarities between the prerecorded and live interactions.

AB - We address the problem of creating believable animations for virtual humans that need to react to the body movements of a human interaction partner in real-time. Our data-driven approach uses prerecorded motion capture data of two interacting persons and performs motion adaptation during the live human-agent interaction. Extending the interaction mesh approach, our main contribution is a new scheme for efficient identification of motions in the prerecorded animation data that are similar to the live interaction. A global low-dimensional posture space serves to select the most similar interaction example, while local, more detail-rich posture spaces are used to identify poses closely matching the human motion. Using the interaction mesh of the selected motion example, an animation can then be synthesized that takes into account both spatial and temporal similarities between the prerecorded and live interactions.

KW - Character animation

KW - interaction mesh

KW - interactive characters

KW - virtual agent

UR - http://www.scopus.com/inward/record.url?scp=84906501922&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84906501922&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-09767-1_57

DO - 10.1007/978-3-319-09767-1_57

M3 - Conference contribution

SN - 9783319097664

VL - 8637 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 463

EP - 476

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -