Pose Estimation of Interacting People using Pictorial Structures

Preben Fihl, Thomas B. Moeslund

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

4 Citations (Scopus)

Abstract

Pose estimation of people have had great progress in recent years but so far research has dealt with single persons. In this paper we address some of the challenges that arise when doing pose estimation of interacting people. We build on the pictorial structures framework and make important contributions by combining color-based appearance and edge information using a measure of the local quality of the appearance feature. In this way we not only combine the two types of features but dynamically find the optimal weighting of them. We further enable the method to handle occlusions by searching a foreground mask for possible occluded body parts and then applying extra strong kinematic constraints to find the true occluded body parts. The effect of applying our two contributions are show through both qualitative and quantitative tests and show a clear improvement on the ability to correctly localize body parts.
Original languageEnglish
Title of host publication7th IEEE International Conference on Advanced Video and Signal-Based Surveillance : for Security, Safety and Monitoring in Smart Environments
Number of pages7
PublisherIEEE Computer Society Press
Publication date2010
Pages462-468
ISBN (Print)978-1-4244-8310-5
DOIs
Publication statusPublished - 2010

Fingerprint

Masks
Kinematics
Color

Cite this

Fihl, P., & Moeslund, T. B. (2010). Pose Estimation of Interacting People using Pictorial Structures. In 7th IEEE International Conference on Advanced Video and Signal-Based Surveillance: for Security, Safety and Monitoring in Smart Environments (pp. 462-468 ). IEEE Computer Society Press. https://doi.org/10.1109/AVSS.2010.27
Fihl, Preben ; Moeslund, Thomas B. / Pose Estimation of Interacting People using Pictorial Structures. 7th IEEE International Conference on Advanced Video and Signal-Based Surveillance: for Security, Safety and Monitoring in Smart Environments. IEEE Computer Society Press, 2010. pp. 462-468
@inproceedings{cf79a8086ed9474aa644af35f28dab81,
title = "Pose Estimation of Interacting People using Pictorial Structures",
abstract = "Pose estimation of people have had great progress in recent years but so far research has dealt with single persons. In this paper we address some of the challenges that arise when doing pose estimation of interacting people. We build on the pictorial structures framework and make important contributions by combining color-based appearance and edge information using a measure of the local quality of the appearance feature. In this way we not only combine the two types of features but dynamically find the optimal weighting of them. We further enable the method to handle occlusions by searching a foreground mask for possible occluded body parts and then applying extra strong kinematic constraints to find the true occluded body parts. The effect of applying our two contributions are show through both qualitative and quantitative tests and show a clear improvement on the ability to correctly localize body parts.",
keywords = "Human pose estimation",
author = "Preben Fihl and Moeslund, {Thomas B.}",
year = "2010",
doi = "10.1109/AVSS.2010.27",
language = "English",
isbn = "978-1-4244-8310-5",
pages = "462--468",
booktitle = "7th IEEE International Conference on Advanced Video and Signal-Based Surveillance",
publisher = "IEEE Computer Society Press",
address = "United States",

}

Fihl, P & Moeslund, TB 2010, Pose Estimation of Interacting People using Pictorial Structures. in 7th IEEE International Conference on Advanced Video and Signal-Based Surveillance: for Security, Safety and Monitoring in Smart Environments. IEEE Computer Society Press, pp. 462-468 . https://doi.org/10.1109/AVSS.2010.27

Pose Estimation of Interacting People using Pictorial Structures. / Fihl, Preben; Moeslund, Thomas B.

7th IEEE International Conference on Advanced Video and Signal-Based Surveillance: for Security, Safety and Monitoring in Smart Environments. IEEE Computer Society Press, 2010. p. 462-468 .

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Pose Estimation of Interacting People using Pictorial Structures

AU - Fihl, Preben

AU - Moeslund, Thomas B.

PY - 2010

Y1 - 2010

N2 - Pose estimation of people have had great progress in recent years but so far research has dealt with single persons. In this paper we address some of the challenges that arise when doing pose estimation of interacting people. We build on the pictorial structures framework and make important contributions by combining color-based appearance and edge information using a measure of the local quality of the appearance feature. In this way we not only combine the two types of features but dynamically find the optimal weighting of them. We further enable the method to handle occlusions by searching a foreground mask for possible occluded body parts and then applying extra strong kinematic constraints to find the true occluded body parts. The effect of applying our two contributions are show through both qualitative and quantitative tests and show a clear improvement on the ability to correctly localize body parts.

AB - Pose estimation of people have had great progress in recent years but so far research has dealt with single persons. In this paper we address some of the challenges that arise when doing pose estimation of interacting people. We build on the pictorial structures framework and make important contributions by combining color-based appearance and edge information using a measure of the local quality of the appearance feature. In this way we not only combine the two types of features but dynamically find the optimal weighting of them. We further enable the method to handle occlusions by searching a foreground mask for possible occluded body parts and then applying extra strong kinematic constraints to find the true occluded body parts. The effect of applying our two contributions are show through both qualitative and quantitative tests and show a clear improvement on the ability to correctly localize body parts.

KW - Human pose estimation

U2 - 10.1109/AVSS.2010.27

DO - 10.1109/AVSS.2010.27

M3 - Article in proceeding

SN - 978-1-4244-8310-5

SP - 462

EP - 468

BT - 7th IEEE International Conference on Advanced Video and Signal-Based Surveillance

PB - IEEE Computer Society Press

ER -

Fihl P, Moeslund TB. Pose Estimation of Interacting People using Pictorial Structures. In 7th IEEE International Conference on Advanced Video and Signal-Based Surveillance: for Security, Safety and Monitoring in Smart Environments. IEEE Computer Society Press. 2010. p. 462-468 https://doi.org/10.1109/AVSS.2010.27