Using scene context to improve action recognition

Juarez Monteiro; Roger Granada; Felipe Meneguzzi; Rodrigo Coelho Barros

doi:10.1007/978-3-030-13469-3_110

Using scene context to improve action recognition

Juarez Monteiro^*, Roger Granada, Felipe Meneguzzi, Rodrigo Coelho Barros

^*Corresponding author for this work

Computing Science

Pontifícia Universidade Católica do Rio Grande do Sul

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

Abstract

Recently action recognition has been used for a variety of applications such as surveillance, smart homes, and in-home elder monitoring. Such applications usually focus on recognizing human actions without taking into account the different scenarios where the action occurs. In this paper, we propose a two-stream architecture that considers not only the movements to identify the action, but also the context scene where the action is performed. Experiments show that the scene context may improve the recognition of certain actions. Our proposed architecture is tested against baselines and the standard two-stream network.

Original language	English
Title of host publication	Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings
Editors	Ruben Vera-Rodriguez, Julian Fierrez, Aythami Morales
Publisher	Springer-Verlag
Pages	954-961
Number of pages	8
ISBN (Print)	9783030134686
DOIs	https://doi.org/10.1007/978-3-030-13469-3_110
Publication status	Published - 2019
Event	23rd Iberoamerican Congress on Pattern Recognition, CIARP 2018 - Madrid, Spain Duration: 19 Nov 2018 → 22 Nov 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11401 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	23rd Iberoamerican Congress on Pattern Recognition, CIARP 2018
Country/Territory	Spain
City	Madrid
Period	19/11/18 → 22/11/18

Bibliographical note

Funding Information:
Acknowledgement. The authors would like to thank CAPES/FAPERGS and Motorola Mobility for partially funding this research. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research.

Publisher Copyright:
© Springer Nature Switzerland AG 2019.

Keywords

Action recognition
Convolutional Neural Networks
Neural networks

Access to Document

10.1007/978-3-030-13469-3_110

Cite this

Monteiro, J., Granada, R., Meneguzzi, F., & Barros, R. C. (2019). Using scene context to improve action recognition. In R. Vera-Rodriguez, J. Fierrez, & A. Morales (Eds.), Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings (pp. 954-961). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11401 LNCS). Springer-Verlag. https://doi.org/10.1007/978-3-030-13469-3_110

Using scene context to improve action recognition. / Monteiro, Juarez; Granada, Roger; Meneguzzi, Felipe et al.
Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings. ed. / Ruben Vera-Rodriguez; Julian Fierrez; Aythami Morales. Springer-Verlag, 2019. p. 954-961 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11401 LNCS).

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

Monteiro, J, Granada, R, Meneguzzi, F & Barros, RC 2019, Using scene context to improve action recognition. in R Vera-Rodriguez, J Fierrez & A Morales (eds), Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11401 LNCS, Springer-Verlag, pp. 954-961, 23rd Iberoamerican Congress on Pattern Recognition, CIARP 2018, Madrid, Spain, 19/11/18. https://doi.org/10.1007/978-3-030-13469-3_110

Monteiro J, Granada R, Meneguzzi F, Barros RC. Using scene context to improve action recognition. In Vera-Rodriguez R, Fierrez J, Morales A, editors, Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings. Springer-Verlag. 2019. p. 954-961. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-13469-3_110

Monteiro, Juarez ; Granada, Roger ; Meneguzzi, Felipe et al. / Using scene context to improve action recognition. Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings. editor / Ruben Vera-Rodriguez ; Julian Fierrez ; Aythami Morales. Springer-Verlag, 2019. pp. 954-961 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{de2de59ff4754d42b7d7d6743affc4f2,

title = "Using scene context to improve action recognition",

abstract = "Recently action recognition has been used for a variety of applications such as surveillance, smart homes, and in-home elder monitoring. Such applications usually focus on recognizing human actions without taking into account the different scenarios where the action occurs. In this paper, we propose a two-stream architecture that considers not only the movements to identify the action, but also the context scene where the action is performed. Experiments show that the scene context may improve the recognition of certain actions. Our proposed architecture is tested against baselines and the standard two-stream network.",

keywords = "Action recognition, Convolutional Neural Networks, Neural networks",

author = "Juarez Monteiro and Roger Granada and Felipe Meneguzzi and Barros, {Rodrigo Coelho}",

note = "Funding Information: Acknowledgement. The authors would like to thank CAPES/FAPERGS and Motorola Mobility for partially funding this research. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research. Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019.; 23rd Iberoamerican Congress on Pattern Recognition, CIARP 2018 ; Conference date: 19-11-2018 Through 22-11-2018",

year = "2019",

doi = "10.1007/978-3-030-13469-3_110",

language = "English",

isbn = "9783030134686",

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

publisher = "Springer-Verlag",

pages = "954--961",

editor = "Ruben Vera-Rodriguez and Julian Fierrez and Aythami Morales",

booktitle = "Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings",

}

TY - GEN

T1 - Using scene context to improve action recognition

AU - Monteiro, Juarez

AU - Granada, Roger

AU - Meneguzzi, Felipe

AU - Barros, Rodrigo Coelho

N1 - Funding Information: Acknowledgement. The authors would like to thank CAPES/FAPERGS and Motorola Mobility for partially funding this research. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research. Publisher Copyright: © Springer Nature Switzerland AG 2019.

PY - 2019

Y1 - 2019

N2 - Recently action recognition has been used for a variety of applications such as surveillance, smart homes, and in-home elder monitoring. Such applications usually focus on recognizing human actions without taking into account the different scenarios where the action occurs. In this paper, we propose a two-stream architecture that considers not only the movements to identify the action, but also the context scene where the action is performed. Experiments show that the scene context may improve the recognition of certain actions. Our proposed architecture is tested against baselines and the standard two-stream network.

AB - Recently action recognition has been used for a variety of applications such as surveillance, smart homes, and in-home elder monitoring. Such applications usually focus on recognizing human actions without taking into account the different scenarios where the action occurs. In this paper, we propose a two-stream architecture that considers not only the movements to identify the action, but also the context scene where the action is performed. Experiments show that the scene context may improve the recognition of certain actions. Our proposed architecture is tested against baselines and the standard two-stream network.

KW - Action recognition

KW - Convolutional Neural Networks

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=85063044065&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-13469-3_110

DO - 10.1007/978-3-030-13469-3_110

M3 - Published conference contribution

AN - SCOPUS:85063044065

SN - 9783030134686

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

SP - 954

EP - 961

BT - Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings

A2 - Vera-Rodriguez, Ruben

A2 - Fierrez, Julian

A2 - Morales, Aythami

PB - Springer-Verlag

T2 - 23rd Iberoamerican Congress on Pattern Recognition, CIARP 2018

Y2 - 19 November 2018 through 22 November 2018

ER -

Using scene context to improve action recognition

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this