Prediction based occluded multitarget tracking using spatio-temporal attention

Heungkyu Lee, June Kim, Hanseok Ko

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper proposes the prediction based occluded multitarget tracking method using spatio-temporal attention mechanism. To cope with occlusion between targets, the proposed method provides an efficient method for more complex analysis by combining object association with partial probability model in spatially attentive window and occlusion activity detection in predicted temporal location. While multiple objects are moving or occluding between them in areas of visual field, a simultaneous tracking of multiple objects tends to fail. This is due to the fact that incompletely estimated feature vectors such as location, color, velocity, and acceleration of a target can provide only ambiguous and missing information. Thus, the spatially and temporally considered mechanism is proposed to track each target before, during, and after occlusion. Robustness of the proposed method is demonstrated with representative simulations.

Original languageEnglish
Pages (from-to)925-938
Number of pages14
JournalInternational Journal of Pattern Recognition and Artificial Intelligence
Volume20
Issue number6
DOIs
Publication statusPublished - 2006 Sep 1

Fingerprint

Color

Keywords

  • Focus of attention
  • Multitarget tracking
  • Occlusion
  • Spatial attention

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Control and Systems Engineering

Cite this

Prediction based occluded multitarget tracking using spatio-temporal attention. / Lee, Heungkyu; Kim, June; Ko, Hanseok.

In: International Journal of Pattern Recognition and Artificial Intelligence, Vol. 20, No. 6, 01.09.2006, p. 925-938.

Research output: Contribution to journalArticle

@article{0600385c51a846f3bfc4b2c75d1a0ea7,
title = "Prediction based occluded multitarget tracking using spatio-temporal attention",
abstract = "This paper proposes the prediction based occluded multitarget tracking method using spatio-temporal attention mechanism. To cope with occlusion between targets, the proposed method provides an efficient method for more complex analysis by combining object association with partial probability model in spatially attentive window and occlusion activity detection in predicted temporal location. While multiple objects are moving or occluding between them in areas of visual field, a simultaneous tracking of multiple objects tends to fail. This is due to the fact that incompletely estimated feature vectors such as location, color, velocity, and acceleration of a target can provide only ambiguous and missing information. Thus, the spatially and temporally considered mechanism is proposed to track each target before, during, and after occlusion. Robustness of the proposed method is demonstrated with representative simulations.",
keywords = "Focus of attention, Multitarget tracking, Occlusion, Spatial attention",
author = "Heungkyu Lee and June Kim and Hanseok Ko",
year = "2006",
month = "9",
day = "1",
doi = "10.1142/S0218001406005058",
language = "English",
volume = "20",
pages = "925--938",
journal = "International Journal of Pattern Recognition and Artificial Intelligence",
issn = "0218-0014",
publisher = "World Scientific Publishing Co. Pte Ltd",
number = "6",

}

TY - JOUR

T1 - Prediction based occluded multitarget tracking using spatio-temporal attention

AU - Lee, Heungkyu

AU - Kim, June

AU - Ko, Hanseok

PY - 2006/9/1

Y1 - 2006/9/1

N2 - This paper proposes the prediction based occluded multitarget tracking method using spatio-temporal attention mechanism. To cope with occlusion between targets, the proposed method provides an efficient method for more complex analysis by combining object association with partial probability model in spatially attentive window and occlusion activity detection in predicted temporal location. While multiple objects are moving or occluding between them in areas of visual field, a simultaneous tracking of multiple objects tends to fail. This is due to the fact that incompletely estimated feature vectors such as location, color, velocity, and acceleration of a target can provide only ambiguous and missing information. Thus, the spatially and temporally considered mechanism is proposed to track each target before, during, and after occlusion. Robustness of the proposed method is demonstrated with representative simulations.

AB - This paper proposes the prediction based occluded multitarget tracking method using spatio-temporal attention mechanism. To cope with occlusion between targets, the proposed method provides an efficient method for more complex analysis by combining object association with partial probability model in spatially attentive window and occlusion activity detection in predicted temporal location. While multiple objects are moving or occluding between them in areas of visual field, a simultaneous tracking of multiple objects tends to fail. This is due to the fact that incompletely estimated feature vectors such as location, color, velocity, and acceleration of a target can provide only ambiguous and missing information. Thus, the spatially and temporally considered mechanism is proposed to track each target before, during, and after occlusion. Robustness of the proposed method is demonstrated with representative simulations.

KW - Focus of attention

KW - Multitarget tracking

KW - Occlusion

KW - Spatial attention

UR - http://www.scopus.com/inward/record.url?scp=33748988484&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33748988484&partnerID=8YFLogxK

U2 - 10.1142/S0218001406005058

DO - 10.1142/S0218001406005058

M3 - Article

AN - SCOPUS:33748988484

VL - 20

SP - 925

EP - 938

JO - International Journal of Pattern Recognition and Artificial Intelligence

JF - International Journal of Pattern Recognition and Artificial Intelligence

SN - 0218-0014

IS - 6

ER -