Physically-based linear blend skinning

Youngbeom Kim, Junghyun Han

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

Abstract

This paper proposes an example-based linear blend skinning system that incorporates the torques applied to the skeletal joints. For this purpose, deformation gradient predictors are trained using a set of examples with the minimum torque and another set with the maximum torque. The run-time algorithm takes an animated skeleton and joint torques as input. It computes the deformation gradients from the predictors. They are then combined with torque-based blending functions. The functions are trained at the preprocessing stage using a set of skeleton-mesh-torque examples with the aid of statistical analysis. The experimental results show that the proposed system runs at real-time and produces quite realistic skinned meshes based on the input torques.

Original languageEnglish
Title of host publicationProceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017
PublisherIADIS
Pages349-350
Number of pages2
ISBN (Electronic)9789898533661
Publication statusPublished - 2017 Jan 1
Event11th International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2017 and International Conference on Big Data Analytics, Data Mining and Computational Intelligence, BigDaCI 2017 - Lisbon, Portugal
Duration: 2017 Jul 212017 Jul 23

Other

Other11th International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2017 and International Conference on Big Data Analytics, Data Mining and Computational Intelligence, BigDaCI 2017
CountryPortugal
CityLisbon
Period17/7/2117/7/23

Fingerprint

Torque
Statistical methods

Keywords

  • Computer animation
  • Linear blend skinning
  • Skeletal joints
  • Torques

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Kim, Y., & Han, J. (2017). Physically-based linear blend skinning. In Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017 (pp. 349-350). IADIS.

Physically-based linear blend skinning. / Kim, Youngbeom; Han, Junghyun.

Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017. IADIS, 2017. p. 349-350.

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

Kim, Y & Han, J 2017, Physically-based linear blend skinning. in Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017. IADIS, pp. 349-350, 11th International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2017 and International Conference on Big Data Analytics, Data Mining and Computational Intelligence, BigDaCI 2017, Lisbon, Portugal, 17/7/21.
Kim Y, Han J. Physically-based linear blend skinning. In Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017. IADIS. 2017. p. 349-350
Kim, Youngbeom ; Han, Junghyun. / Physically-based linear blend skinning. Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017. IADIS, 2017. pp. 349-350
@inproceedings{89e83b8d754f42a386116e8564ccf327,
title = "Physically-based linear blend skinning",
abstract = "This paper proposes an example-based linear blend skinning system that incorporates the torques applied to the skeletal joints. For this purpose, deformation gradient predictors are trained using a set of examples with the minimum torque and another set with the maximum torque. The run-time algorithm takes an animated skeleton and joint torques as input. It computes the deformation gradients from the predictors. They are then combined with torque-based blending functions. The functions are trained at the preprocessing stage using a set of skeleton-mesh-torque examples with the aid of statistical analysis. The experimental results show that the proposed system runs at real-time and produces quite realistic skinned meshes based on the input torques.",
keywords = "Computer animation, Linear blend skinning, Skeletal joints, Torques",
author = "Youngbeom Kim and Junghyun Han",
year = "2017",
month = "1",
day = "1",
language = "English",
pages = "349--350",
booktitle = "Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017",
publisher = "IADIS",

}

TY - GEN

T1 - Physically-based linear blend skinning

AU - Kim, Youngbeom

AU - Han, Junghyun

PY - 2017/1/1

Y1 - 2017/1/1

N2 - This paper proposes an example-based linear blend skinning system that incorporates the torques applied to the skeletal joints. For this purpose, deformation gradient predictors are trained using a set of examples with the minimum torque and another set with the maximum torque. The run-time algorithm takes an animated skeleton and joint torques as input. It computes the deformation gradients from the predictors. They are then combined with torque-based blending functions. The functions are trained at the preprocessing stage using a set of skeleton-mesh-torque examples with the aid of statistical analysis. The experimental results show that the proposed system runs at real-time and produces quite realistic skinned meshes based on the input torques.

AB - This paper proposes an example-based linear blend skinning system that incorporates the torques applied to the skeletal joints. For this purpose, deformation gradient predictors are trained using a set of examples with the minimum torque and another set with the maximum torque. The run-time algorithm takes an animated skeleton and joint torques as input. It computes the deformation gradients from the predictors. They are then combined with torque-based blending functions. The functions are trained at the preprocessing stage using a set of skeleton-mesh-torque examples with the aid of statistical analysis. The experimental results show that the proposed system runs at real-time and produces quite realistic skinned meshes based on the input torques.

KW - Computer animation

KW - Linear blend skinning

KW - Skeletal joints

KW - Torques

UR - http://www.scopus.com/inward/record.url?scp=85040169941&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040169941&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85040169941

SP - 349

EP - 350

BT - Proceedings of the International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing 2017 and Big Data Analytics, Data Mining and Computational Intelligence 2017 - Part of the Multi Conference on Computer Science and Information Systems 2017

PB - IADIS

ER -