Single Image, Context Aware Action Estimation in Sports

Christian Lanius, Daisuke Kobayashi, Kazushige Ouchi, Yoshimitsu Aoki

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

Abstract

Sports video analysis enables professional teams to prepare and practice more efficiently than ever before. It allows the team to work on their weak points and analyze upcoming opponents' strategies. As segmentation is a well researched topic, we assume a sports video where each frame is segmented into actors. We show that the location of the actors is correlated to their action. We develop a model to predict each actor's action in a context consistent and location aware manner. The set of context actors is an unordered set, thus we propose a permutation invariant structure for our classifier. We evaluate a location and a context aware network for single image sports action recognition on two datasets. One is an existing dataset of volleyball games. We present a new dataset for actor recognition in rugby. We compare the results to using only the image of the actor.

Original languageEnglish
Title of host publicationProceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018
EditorsRichard Chbeir, Gabriella Sanniti di Baja, Luigi Gallo, Kokou Yetongnon, Albert Dipanda, Modesto Castrillon-Santana
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages664-671
Number of pages8
ISBN (Electronic)9781538693858
DOIs
Publication statusPublished - 2019 May 3
Event14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018 - Las Palmas de Gran Canaria, Spain
Duration: 2018 Nov 262018 Nov 29

Publication series

NameProceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018

Conference

Conference14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018
CountrySpain
CityLas Palmas de Gran Canaria
Period18/11/2618/11/29

Fingerprint

Sports
Classifiers

Keywords

  • Context
  • Deep-learning
  • Sports action recognition

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition

Cite this

Lanius, C., Kobayashi, D., Ouchi, K., & Aoki, Y. (2019). Single Image, Context Aware Action Estimation in Sports. In R. Chbeir, G. S. di Baja, L. Gallo, K. Yetongnon, A. Dipanda, & M. Castrillon-Santana (Eds.), Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018 (pp. 664-671). [8706211] (Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SITIS.2018.00107

Single Image, Context Aware Action Estimation in Sports. / Lanius, Christian; Kobayashi, Daisuke; Ouchi, Kazushige; Aoki, Yoshimitsu.

Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018. ed. / Richard Chbeir; Gabriella Sanniti di Baja; Luigi Gallo; Kokou Yetongnon; Albert Dipanda; Modesto Castrillon-Santana. Institute of Electrical and Electronics Engineers Inc., 2019. p. 664-671 8706211 (Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018).

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

Lanius, C, Kobayashi, D, Ouchi, K & Aoki, Y 2019, Single Image, Context Aware Action Estimation in Sports. in R Chbeir, GS di Baja, L Gallo, K Yetongnon, A Dipanda & M Castrillon-Santana (eds), Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018., 8706211, Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018, Institute of Electrical and Electronics Engineers Inc., pp. 664-671, 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018, Las Palmas de Gran Canaria, Spain, 18/11/26. https://doi.org/10.1109/SITIS.2018.00107
Lanius C, Kobayashi D, Ouchi K, Aoki Y. Single Image, Context Aware Action Estimation in Sports. In Chbeir R, di Baja GS, Gallo L, Yetongnon K, Dipanda A, Castrillon-Santana M, editors, Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018. Institute of Electrical and Electronics Engineers Inc. 2019. p. 664-671. 8706211. (Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018). https://doi.org/10.1109/SITIS.2018.00107
Lanius, Christian ; Kobayashi, Daisuke ; Ouchi, Kazushige ; Aoki, Yoshimitsu. / Single Image, Context Aware Action Estimation in Sports. Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018. editor / Richard Chbeir ; Gabriella Sanniti di Baja ; Luigi Gallo ; Kokou Yetongnon ; Albert Dipanda ; Modesto Castrillon-Santana. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 664-671 (Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018).
@inproceedings{aeed1911388a47c8af06d1db48f3f29e,
title = "Single Image, Context Aware Action Estimation in Sports",
abstract = "Sports video analysis enables professional teams to prepare and practice more efficiently than ever before. It allows the team to work on their weak points and analyze upcoming opponents' strategies. As segmentation is a well researched topic, we assume a sports video where each frame is segmented into actors. We show that the location of the actors is correlated to their action. We develop a model to predict each actor's action in a context consistent and location aware manner. The set of context actors is an unordered set, thus we propose a permutation invariant structure for our classifier. We evaluate a location and a context aware network for single image sports action recognition on two datasets. One is an existing dataset of volleyball games. We present a new dataset for actor recognition in rugby. We compare the results to using only the image of the actor.",
keywords = "Context, Deep-learning, Sports action recognition",
author = "Christian Lanius and Daisuke Kobayashi and Kazushige Ouchi and Yoshimitsu Aoki",
year = "2019",
month = "5",
day = "3",
doi = "10.1109/SITIS.2018.00107",
language = "English",
series = "Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "664--671",
editor = "Richard Chbeir and {di Baja}, {Gabriella Sanniti} and Luigi Gallo and Kokou Yetongnon and Albert Dipanda and Modesto Castrillon-Santana",
booktitle = "Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018",

}

TY - GEN

T1 - Single Image, Context Aware Action Estimation in Sports

AU - Lanius, Christian

AU - Kobayashi, Daisuke

AU - Ouchi, Kazushige

AU - Aoki, Yoshimitsu

PY - 2019/5/3

Y1 - 2019/5/3

N2 - Sports video analysis enables professional teams to prepare and practice more efficiently than ever before. It allows the team to work on their weak points and analyze upcoming opponents' strategies. As segmentation is a well researched topic, we assume a sports video where each frame is segmented into actors. We show that the location of the actors is correlated to their action. We develop a model to predict each actor's action in a context consistent and location aware manner. The set of context actors is an unordered set, thus we propose a permutation invariant structure for our classifier. We evaluate a location and a context aware network for single image sports action recognition on two datasets. One is an existing dataset of volleyball games. We present a new dataset for actor recognition in rugby. We compare the results to using only the image of the actor.

AB - Sports video analysis enables professional teams to prepare and practice more efficiently than ever before. It allows the team to work on their weak points and analyze upcoming opponents' strategies. As segmentation is a well researched topic, we assume a sports video where each frame is segmented into actors. We show that the location of the actors is correlated to their action. We develop a model to predict each actor's action in a context consistent and location aware manner. The set of context actors is an unordered set, thus we propose a permutation invariant structure for our classifier. We evaluate a location and a context aware network for single image sports action recognition on two datasets. One is an existing dataset of volleyball games. We present a new dataset for actor recognition in rugby. We compare the results to using only the image of the actor.

KW - Context

KW - Deep-learning

KW - Sports action recognition

UR - http://www.scopus.com/inward/record.url?scp=85065923028&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065923028&partnerID=8YFLogxK

U2 - 10.1109/SITIS.2018.00107

DO - 10.1109/SITIS.2018.00107

M3 - Conference contribution

AN - SCOPUS:85065923028

T3 - Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018

SP - 664

EP - 671

BT - Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018

A2 - Chbeir, Richard

A2 - di Baja, Gabriella Sanniti

A2 - Gallo, Luigi

A2 - Yetongnon, Kokou

A2 - Dipanda, Albert

A2 - Castrillon-Santana, Modesto

PB - Institute of Electrical and Electronics Engineers Inc.

ER -