Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks

Minho Shim; Young Hwi Kim; Kyungmin Kim; Seon Joo Kim

doi:10.1007/978-3-030-01267-0_25

Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks

Minho Shim, Young Hwi Kim, Kyungmin Kim, Seon Joo Kim

College of Computing

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Citations (Scopus)

Abstract

A major obstacle in teaching machines to understand videos is the lack of training data, as creating temporal annotations for long videos requires a huge amount of human effort. To this end, we introduce a new large-scale baseball video dataset called the BBDB, which is produced semi-automatically by using play-by-play texts available online. The BBDB contains 4200+hr of baseball game videos with 400k+ temporally annotated activity segments. The new dataset has several major challenging factors compared to other datasets: (1) the dataset contains a large number of visually similar segments with different labels. (2) It can be used for many video understanding tasks including video recognition, localization, text-video alignment, video highlight generation, and data imbalance problem. To observe the potential of the BBDB, we conducted extensive experiments by running many different types of video understanding algorithms on our new dataset. The database is available at https://sites.google.com/site/eccv2018bbdb/.

Original language	English
Title of host publication	Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
Editors	Yair Weiss, Vittorio Ferrari, Cristian Sminchisescu, Martial Hebert
Publisher	Springer Verlag
Pages	420-437
Number of pages	18
ISBN (Print)	9783030012663
DOIs	https://doi.org/10.1007/978-3-030-01267-0_25
Publication status	Published - 2018
Event	15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany Duration: 2018 Sept 8 → 2018 Sept 14

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11219 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	15th European Conference on Computer Vision, ECCV 2018
Country/Territory	Germany
City	Munich
Period	18/9/8 → 18/9/14

Bibliographical note

Funding Information:
Acknowledgment. This work was supported by Samsung Research Funding Center of Samsung Electronics under Project Number SRFC-IT1701-01.

Publisher Copyright:
© Springer Nature Switzerland AG 2018.

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-030-01267-0_25

Cite this

Shim, M., Kim, Y. H., Kim, K., & Kim, S. J. (2018). Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks. In Y. Weiss, V. Ferrari, C. Sminchisescu, & M. Hebert (Eds.), Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings (pp. 420-437). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11219 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-01267-0_25

Shim, Minho ; Kim, Young Hwi ; Kim, Kyungmin et al. / Teaching machines to understand baseball games : Large-scale baseball video database for multiple video understanding tasks. Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. editor / Yair Weiss ; Vittorio Ferrari ; Cristian Sminchisescu ; Martial Hebert. Springer Verlag, 2018. pp. 420-437 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{1c3a7ad419c7498d935c0640421d4a5d,

title = "Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks",

abstract = "A major obstacle in teaching machines to understand videos is the lack of training data, as creating temporal annotations for long videos requires a huge amount of human effort. To this end, we introduce a new large-scale baseball video dataset called the BBDB, which is produced semi-automatically by using play-by-play texts available online. The BBDB contains 4200+hr of baseball game videos with 400k+ temporally annotated activity segments. The new dataset has several major challenging factors compared to other datasets: (1) the dataset contains a large number of visually similar segments with different labels. (2) It can be used for many video understanding tasks including video recognition, localization, text-video alignment, video highlight generation, and data imbalance problem. To observe the potential of the BBDB, we conducted extensive experiments by running many different types of video understanding algorithms on our new dataset. The database is available at https://sites.google.com/site/eccv2018bbdb/.",

author = "Minho Shim and Kim, {Young Hwi} and Kyungmin Kim and Kim, {Seon Joo}",

note = "Funding Information: Acknowledgment. This work was supported by Samsung Research Funding Center of Samsung Electronics under Project Number SRFC-IT1701-01. Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2018.; 15th European Conference on Computer Vision, ECCV 2018 ; Conference date: 08-09-2018 Through 14-09-2018",

year = "2018",

doi = "10.1007/978-3-030-01267-0_25",

language = "English",

isbn = "9783030012663",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "420--437",

editor = "Yair Weiss and Vittorio Ferrari and Cristian Sminchisescu and Martial Hebert",

booktitle = "Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings",

address = "Germany",

}

Shim, M, Kim, YH, Kim, K & Kim, SJ 2018, Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks. in Y Weiss, V Ferrari, C Sminchisescu & M Hebert (eds), Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11219 LNCS, Springer Verlag, pp. 420-437, 15th European Conference on Computer Vision, ECCV 2018, Munich, Germany, 18/9/8. https://doi.org/10.1007/978-3-030-01267-0_25

Teaching machines to understand baseball games : Large-scale baseball video database for multiple video understanding tasks. / Shim, Minho; Kim, Young Hwi; Kim, Kyungmin et al.

Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. ed. / Yair Weiss; Vittorio Ferrari; Cristian Sminchisescu; Martial Hebert. Springer Verlag, 2018. p. 420-437 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11219 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Teaching machines to understand baseball games

T2 - 15th European Conference on Computer Vision, ECCV 2018

AU - Shim, Minho

AU - Kim, Young Hwi

AU - Kim, Kyungmin

AU - Kim, Seon Joo

N1 - Funding Information: Acknowledgment. This work was supported by Samsung Research Funding Center of Samsung Electronics under Project Number SRFC-IT1701-01. Publisher Copyright: © Springer Nature Switzerland AG 2018.

PY - 2018

Y1 - 2018

N2 - A major obstacle in teaching machines to understand videos is the lack of training data, as creating temporal annotations for long videos requires a huge amount of human effort. To this end, we introduce a new large-scale baseball video dataset called the BBDB, which is produced semi-automatically by using play-by-play texts available online. The BBDB contains 4200+hr of baseball game videos with 400k+ temporally annotated activity segments. The new dataset has several major challenging factors compared to other datasets: (1) the dataset contains a large number of visually similar segments with different labels. (2) It can be used for many video understanding tasks including video recognition, localization, text-video alignment, video highlight generation, and data imbalance problem. To observe the potential of the BBDB, we conducted extensive experiments by running many different types of video understanding algorithms on our new dataset. The database is available at https://sites.google.com/site/eccv2018bbdb/.

AB - A major obstacle in teaching machines to understand videos is the lack of training data, as creating temporal annotations for long videos requires a huge amount of human effort. To this end, we introduce a new large-scale baseball video dataset called the BBDB, which is produced semi-automatically by using play-by-play texts available online. The BBDB contains 4200+hr of baseball game videos with 400k+ temporally annotated activity segments. The new dataset has several major challenging factors compared to other datasets: (1) the dataset contains a large number of visually similar segments with different labels. (2) It can be used for many video understanding tasks including video recognition, localization, text-video alignment, video highlight generation, and data imbalance problem. To observe the potential of the BBDB, we conducted extensive experiments by running many different types of video understanding algorithms on our new dataset. The database is available at https://sites.google.com/site/eccv2018bbdb/.

UR - http://www.scopus.com/inward/record.url?scp=85055442066&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85055442066&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-01267-0_25

DO - 10.1007/978-3-030-01267-0_25

M3 - Conference contribution

AN - SCOPUS:85055442066

SN - 9783030012663

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

SP - 420

EP - 437

BT - Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings

A2 - Weiss, Yair

A2 - Ferrari, Vittorio

A2 - Sminchisescu, Cristian

A2 - Hebert, Martial

PB - Springer Verlag

Y2 - 8 September 2018 through 14 September 2018

ER -

Shim M, Kim YH, Kim K, Kim SJ. Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks. In Weiss Y, Ferrari V, Sminchisescu C, Hebert M, editors, Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. Springer Verlag. 2018. p. 420-437. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-01267-0_25

Teaching machines to understand baseball games: Large-scale baseball video database for multiple video understanding tasks

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this