Action recognition in the dark via deep representation learning

Research output: Book chapter/Published conference paperConference paper

Abstract

Human action recognition for automated video surveillance applications is an interesting but a daunting task especially if the videos are captured in unfavourable lighting conditions. These situations encourage the use of multi-sensor video streams. However, simultaneous activity recognition from multiple video streams is a difficult problem due to their complementary and noisy nature. This paper proposes simultaneous action recognition from multiple video streams using deep multi-view representation learning. Furthermore, it introduces a spatio-temporal feature based correlation filter, for simultaneous detection and recognition of multiple human actions in low-light conditions. We evaluated the performance of our proposed filter with extensive experimentation on nighttime action datasets. Experimental results indicate the effectiveness of deep fusion scheme for robust action recognition in extremely low-light conditions.
Original languageEnglish
Title of host publicationIEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages131-136
Number of pages6
ISBN (Electronic)9781728102474
DOIs
Publication statusPublished - 06 May 2019
Event3rd IEEE International Conference on Image Processing, Applications and Systems, IPAS 2018 - Sophia Antipolis, France
Duration: 12 Dec 201814 Dec 2018

Conference

Conference3rd IEEE International Conference on Image Processing, Applications and Systems, IPAS 2018
CountryFrance
CitySophia Antipolis
Period12/12/1814/12/18

Fingerprint

Fusion reactions
Lighting
Sensors

Cite this

Ulhaq, A. (2019). Action recognition in the dark via deep representation learning. In IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018 (pp. 131-136). [8708903] IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/IPAS.2018.8708903
Ulhaq, Anwaar. / Action recognition in the dark via deep representation learning. IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018. IEEE, Institute of Electrical and Electronics Engineers, 2019. pp. 131-136
@inproceedings{054dcf4eb39e4dd4a02c18d946c27f54,
title = "Action recognition in the dark via deep representation learning",
abstract = "Human action recognition for automated video surveillance applications is an interesting but a daunting task especially if the videos are captured in unfavourable lighting conditions. These situations encourage the use of multi-sensor video streams. However, simultaneous activity recognition from multiple video streams is a difficult problem due to their complementary and noisy nature. This paper proposes simultaneous action recognition from multiple video streams using deep multi-view representation learning. Furthermore, it introduces a spatio-temporal feature based correlation filter, for simultaneous detection and recognition of multiple human actions in low-light conditions. We evaluated the performance of our proposed filter with extensive experimentation on nighttime action datasets. Experimental results indicate the effectiveness of deep fusion scheme for robust action recognition in extremely low-light conditions.",
author = "Anwaar Ulhaq",
year = "2019",
month = "5",
day = "6",
doi = "10.1109/IPAS.2018.8708903",
language = "English",
pages = "131--136",
booktitle = "IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States",

}

Ulhaq, A 2019, Action recognition in the dark via deep representation learning. in IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018., 8708903, IEEE, Institute of Electrical and Electronics Engineers, pp. 131-136, 3rd IEEE International Conference on Image Processing, Applications and Systems, IPAS 2018, Sophia Antipolis, France, 12/12/18. https://doi.org/10.1109/IPAS.2018.8708903

Action recognition in the dark via deep representation learning. / Ulhaq, Anwaar.

IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018. IEEE, Institute of Electrical and Electronics Engineers, 2019. p. 131-136 8708903.

Research output: Book chapter/Published conference paperConference paper

TY - GEN

T1 - Action recognition in the dark via deep representation learning

AU - Ulhaq, Anwaar

PY - 2019/5/6

Y1 - 2019/5/6

N2 - Human action recognition for automated video surveillance applications is an interesting but a daunting task especially if the videos are captured in unfavourable lighting conditions. These situations encourage the use of multi-sensor video streams. However, simultaneous activity recognition from multiple video streams is a difficult problem due to their complementary and noisy nature. This paper proposes simultaneous action recognition from multiple video streams using deep multi-view representation learning. Furthermore, it introduces a spatio-temporal feature based correlation filter, for simultaneous detection and recognition of multiple human actions in low-light conditions. We evaluated the performance of our proposed filter with extensive experimentation on nighttime action datasets. Experimental results indicate the effectiveness of deep fusion scheme for robust action recognition in extremely low-light conditions.

AB - Human action recognition for automated video surveillance applications is an interesting but a daunting task especially if the videos are captured in unfavourable lighting conditions. These situations encourage the use of multi-sensor video streams. However, simultaneous activity recognition from multiple video streams is a difficult problem due to their complementary and noisy nature. This paper proposes simultaneous action recognition from multiple video streams using deep multi-view representation learning. Furthermore, it introduces a spatio-temporal feature based correlation filter, for simultaneous detection and recognition of multiple human actions in low-light conditions. We evaluated the performance of our proposed filter with extensive experimentation on nighttime action datasets. Experimental results indicate the effectiveness of deep fusion scheme for robust action recognition in extremely low-light conditions.

UR - http://www.scopus.com/inward/record.url?scp=85066305106&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066305106&partnerID=8YFLogxK

U2 - 10.1109/IPAS.2018.8708903

DO - 10.1109/IPAS.2018.8708903

M3 - Conference paper

SP - 131

EP - 136

BT - IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018

PB - IEEE, Institute of Electrical and Electronics Engineers

ER -

Ulhaq A. Action recognition in the dark via deep representation learning. In IEEE 3rd International Conference on Image Processing, Applications and Systems, IPAS 2018. IEEE, Institute of Electrical and Electronics Engineers. 2019. p. 131-136. 8708903 https://doi.org/10.1109/IPAS.2018.8708903