Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data

Jürgen Bernard, Eduard Dobermann, Anna Vögele, Björn Krüger, Jörn Kohlhammer, Dieter W. Fellner

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

The characterization and abstraction of large multivariate time series data often poses challenges with respect to effectiveness or efficiency. Using the example of human motion capture data challenges exist in creating compact solutions that still reflect semantics and kinematics in a meaningful way. We present a visual-interactive approach for the semi-supervised labeling of human motion capture data. Users are enabled to assign labels to the data which can subsequently be used to represent the multivariate time series as sequences of motion classes. The approach combines multiple views supporting the user in the visual-interactive labeling process. Visual guidance concepts further ease the labeling process by propagating the results of supportive algorithmic models. The abstraction of motion capture data to sequences of event intervals allows overview and detail-on-demand visualizations even for large and heterogeneous data collections. The guided selection of candidate data for the extension and improvement of the labeling closes the feedback loop of the semi-supervised workflow. We demonstrate the effectiveness and the efficiency of the approach in two usage scenarios, taking visual-interactive learning and human motion synthesis as examples.
Original languageUndefined/Unknown
Title of host publicationVisualization and Data Analysis (VDA)
PublisherSociety for Imaging Science and Technology
Pages34-45
Number of pages12
DOIs
Publication statusPublished - 2017

Publication series

NameElectronic Imaging
PublisherIST, Springfield

Fields of Expertise

  • Information, Communication & Computing

Cite this

Bernard, J., Dobermann, E., Vögele, A., Krüger, B., Kohlhammer, J., & Fellner, D. W. (2017). Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data. In Visualization and Data Analysis (VDA) (pp. 34-45). (Electronic Imaging). Society for Imaging Science and Technology. https://doi.org/10.2352/ISSN.2470-1173.2017.1.VDA-387

Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data. / Bernard, Jürgen; Dobermann, Eduard; Vögele, Anna; Krüger, Björn; Kohlhammer, Jörn; Fellner, Dieter W.

Visualization and Data Analysis (VDA). Society for Imaging Science and Technology, 2017. p. 34-45 (Electronic Imaging).

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Bernard, J, Dobermann, E, Vögele, A, Krüger, B, Kohlhammer, J & Fellner, DW 2017, Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data. in Visualization and Data Analysis (VDA). Electronic Imaging, Society for Imaging Science and Technology, pp. 34-45. https://doi.org/10.2352/ISSN.2470-1173.2017.1.VDA-387
Bernard J, Dobermann E, Vögele A, Krüger B, Kohlhammer J, Fellner DW. Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data. In Visualization and Data Analysis (VDA). Society for Imaging Science and Technology. 2017. p. 34-45. (Electronic Imaging). https://doi.org/10.2352/ISSN.2470-1173.2017.1.VDA-387
Bernard, Jürgen ; Dobermann, Eduard ; Vögele, Anna ; Krüger, Björn ; Kohlhammer, Jörn ; Fellner, Dieter W. / Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data. Visualization and Data Analysis (VDA). Society for Imaging Science and Technology, 2017. pp. 34-45 (Electronic Imaging).
@inproceedings{865984bb538949fab79dddc2695bf1c9,
title = "Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data",
abstract = "The characterization and abstraction of large multivariate time series data often poses challenges with respect to effectiveness or efficiency. Using the example of human motion capture data challenges exist in creating compact solutions that still reflect semantics and kinematics in a meaningful way. We present a visual-interactive approach for the semi-supervised labeling of human motion capture data. Users are enabled to assign labels to the data which can subsequently be used to represent the multivariate time series as sequences of motion classes. The approach combines multiple views supporting the user in the visual-interactive labeling process. Visual guidance concepts further ease the labeling process by propagating the results of supportive algorithmic models. The abstraction of motion capture data to sequences of event intervals allows overview and detail-on-demand visualizations even for large and heterogeneous data collections. The guided selection of candidate data for the extension and improvement of the labeling closes the feedback loop of the semi-supervised workflow. We demonstrate the effectiveness and the efficiency of the approach in two usage scenarios, taking visual-interactive learning and human motion synthesis as examples.",
author = "J{\"u}rgen Bernard and Eduard Dobermann and Anna V{\"o}gele and Bj{\"o}rn Kr{\"u}ger and J{\"o}rn Kohlhammer and Fellner, {Dieter W.}",
year = "2017",
doi = "10.2352/ISSN.2470-1173.2017.1.VDA-387",
language = "undefiniert/unbekannt",
series = "Electronic Imaging",
publisher = "Society for Imaging Science and Technology",
pages = "34--45",
booktitle = "Visualization and Data Analysis (VDA)",
address = "USA / Vereinigte Staaten",

}

TY - GEN

T1 - Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data

AU - Bernard, Jürgen

AU - Dobermann, Eduard

AU - Vögele, Anna

AU - Krüger, Björn

AU - Kohlhammer, Jörn

AU - Fellner, Dieter W.

PY - 2017

Y1 - 2017

N2 - The characterization and abstraction of large multivariate time series data often poses challenges with respect to effectiveness or efficiency. Using the example of human motion capture data challenges exist in creating compact solutions that still reflect semantics and kinematics in a meaningful way. We present a visual-interactive approach for the semi-supervised labeling of human motion capture data. Users are enabled to assign labels to the data which can subsequently be used to represent the multivariate time series as sequences of motion classes. The approach combines multiple views supporting the user in the visual-interactive labeling process. Visual guidance concepts further ease the labeling process by propagating the results of supportive algorithmic models. The abstraction of motion capture data to sequences of event intervals allows overview and detail-on-demand visualizations even for large and heterogeneous data collections. The guided selection of candidate data for the extension and improvement of the labeling closes the feedback loop of the semi-supervised workflow. We demonstrate the effectiveness and the efficiency of the approach in two usage scenarios, taking visual-interactive learning and human motion synthesis as examples.

AB - The characterization and abstraction of large multivariate time series data often poses challenges with respect to effectiveness or efficiency. Using the example of human motion capture data challenges exist in creating compact solutions that still reflect semantics and kinematics in a meaningful way. We present a visual-interactive approach for the semi-supervised labeling of human motion capture data. Users are enabled to assign labels to the data which can subsequently be used to represent the multivariate time series as sequences of motion classes. The approach combines multiple views supporting the user in the visual-interactive labeling process. Visual guidance concepts further ease the labeling process by propagating the results of supportive algorithmic models. The abstraction of motion capture data to sequences of event intervals allows overview and detail-on-demand visualizations even for large and heterogeneous data collections. The guided selection of candidate data for the extension and improvement of the labeling closes the feedback loop of the semi-supervised workflow. We demonstrate the effectiveness and the efficiency of the approach in two usage scenarios, taking visual-interactive learning and human motion synthesis as examples.

U2 - 10.2352/ISSN.2470-1173.2017.1.VDA-387

DO - 10.2352/ISSN.2470-1173.2017.1.VDA-387

M3 - Beitrag in einem Konferenzband

T3 - Electronic Imaging

SP - 34

EP - 45

BT - Visualization and Data Analysis (VDA)

PB - Society for Imaging Science and Technology

ER -