Online training of object detectors from unlabeled surveillance video

Hasan Çelik, Alan Hanjalic, Emile A. Hendriks, Sabri Boughorbel

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

8 Citations (Scopus)

Abstract

One of the decisive steps in automated surveillance and monitoring is object detection. A standard approach to constructing object detectors consists of annotating large data sets and using them to train a detector. Nevertheless, due to unavoidable constraints of a typical training data set, supervised approaches are inappropriate for building generic systems applicable to a wide diversity of camera setups and scenes. To make a step towards a more generic solution, we propose in this paper a method capable of learning and detecting, in an online and unsupervised setup, the dominant object class in a general scene. The effectiveness of our method is experimentally demonstrated on four representative video sequences.

Original languageEnglish
Title of host publication2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops
DOIs
Publication statusPublished - 2008
Externally publishedYes
Event2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops - Anchorage, AK, United States
Duration: 23 Jun 200828 Jun 2008

Other

Other2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops
CountryUnited States
CityAnchorage, AK
Period23/6/0828/6/08

Fingerprint

Detectors
Cameras
Monitoring
Object detection

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering

Cite this

Çelik, H., Hanjalic, A., Hendriks, E. A., & Boughorbel, S. (2008). Online training of object detectors from unlabeled surveillance video. In 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops [4563067] https://doi.org/10.1109/CVPRW.2008.4563067

Online training of object detectors from unlabeled surveillance video. / Çelik, Hasan; Hanjalic, Alan; Hendriks, Emile A.; Boughorbel, Sabri.

2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops. 2008. 4563067.

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

Çelik, H, Hanjalic, A, Hendriks, EA & Boughorbel, S 2008, Online training of object detectors from unlabeled surveillance video. in 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops., 4563067, 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops, Anchorage, AK, United States, 23/6/08. https://doi.org/10.1109/CVPRW.2008.4563067
Çelik H, Hanjalic A, Hendriks EA, Boughorbel S. Online training of object detectors from unlabeled surveillance video. In 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops. 2008. 4563067 https://doi.org/10.1109/CVPRW.2008.4563067
Çelik, Hasan ; Hanjalic, Alan ; Hendriks, Emile A. ; Boughorbel, Sabri. / Online training of object detectors from unlabeled surveillance video. 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops. 2008.
@inproceedings{2043d1b3d837433780ff3c67b17e5eba,
title = "Online training of object detectors from unlabeled surveillance video",
abstract = "One of the decisive steps in automated surveillance and monitoring is object detection. A standard approach to constructing object detectors consists of annotating large data sets and using them to train a detector. Nevertheless, due to unavoidable constraints of a typical training data set, supervised approaches are inappropriate for building generic systems applicable to a wide diversity of camera setups and scenes. To make a step towards a more generic solution, we propose in this paper a method capable of learning and detecting, in an online and unsupervised setup, the dominant object class in a general scene. The effectiveness of our method is experimentally demonstrated on four representative video sequences.",
author = "Hasan {\cC}elik and Alan Hanjalic and Hendriks, {Emile A.} and Sabri Boughorbel",
year = "2008",
doi = "10.1109/CVPRW.2008.4563067",
language = "English",
isbn = "9781424423408",
booktitle = "2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops",

}

TY - GEN

T1 - Online training of object detectors from unlabeled surveillance video

AU - Çelik, Hasan

AU - Hanjalic, Alan

AU - Hendriks, Emile A.

AU - Boughorbel, Sabri

PY - 2008

Y1 - 2008

N2 - One of the decisive steps in automated surveillance and monitoring is object detection. A standard approach to constructing object detectors consists of annotating large data sets and using them to train a detector. Nevertheless, due to unavoidable constraints of a typical training data set, supervised approaches are inappropriate for building generic systems applicable to a wide diversity of camera setups and scenes. To make a step towards a more generic solution, we propose in this paper a method capable of learning and detecting, in an online and unsupervised setup, the dominant object class in a general scene. The effectiveness of our method is experimentally demonstrated on four representative video sequences.

AB - One of the decisive steps in automated surveillance and monitoring is object detection. A standard approach to constructing object detectors consists of annotating large data sets and using them to train a detector. Nevertheless, due to unavoidable constraints of a typical training data set, supervised approaches are inappropriate for building generic systems applicable to a wide diversity of camera setups and scenes. To make a step towards a more generic solution, we propose in this paper a method capable of learning and detecting, in an online and unsupervised setup, the dominant object class in a general scene. The effectiveness of our method is experimentally demonstrated on four representative video sequences.

UR - http://www.scopus.com/inward/record.url?scp=51849132416&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=51849132416&partnerID=8YFLogxK

U2 - 10.1109/CVPRW.2008.4563067

DO - 10.1109/CVPRW.2008.4563067

M3 - Conference contribution

AN - SCOPUS:51849132416

SN - 9781424423408

BT - 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops

ER -