Decontaminate feature for tracking: adaptive tracking via evolutionary feature subset

Qiaoyuan Liu, Yuru Wang, Minghao Yin, Jinchang Ren, Ruizhi Li

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Although various visual tracking algorithms have been proposed in the last 2-3 decades, it remains a challenging problem for effective tracking with fast motion, deformation, occlusion et al. Under complex tracking conditions, most tracking models are not discriminative and adaptive enough. When the combined feature vectors are inputted to the visual models, this may lead to redundancy caused low efficiency and ambiguity caused poor performance. In this paper, an effective tracking algorithm is proposed to decontaminate features for each video sequence adaptively, where the visual modeling is treated as an optimization problem from the perspective of evolution. Every feature vector is compared to a biological individual and then decontaminated via classical evolutionary algorithms. With the optimized subsets of features, “Curse of Dimensionality” has been avoided whilst the accuracy of the visual model has been improved. The proposed algorithm has been tested on several publicly available datasets with various tracking challenges and benchmarked with a number of state-of-the-art approaches. The comprehensive experiments have demonstrated the efficacy of the proposed methodology.
LanguageEnglish
Article number063025
Number of pages10
JournalJournal of Electronic Imaging
Volume26
Issue number6
Early online date12 Dec 2017
DOIs
Publication statusE-pub ahead of print - 12 Dec 2017

Fingerprint

Set theory
set theory
Evolutionary algorithms
Redundancy
optical tracking
occlusion
redundancy
ambiguity
methodology
Experiments
optimization

Keywords

  • visual tracking
  • feature subset
  • decontaminate
  • evolutionary algorithm

Cite this

Liu, Qiaoyuan ; Wang, Yuru ; Yin, Minghao ; Ren, Jinchang ; Li, Ruizhi. / Decontaminate feature for tracking : adaptive tracking via evolutionary feature subset. In: Journal of Electronic Imaging. 2017 ; Vol. 26, No. 6.
@article{dfd31224c5d7476da5bf8c81401cabf9,
title = "Decontaminate feature for tracking: adaptive tracking via evolutionary feature subset",
abstract = "Although various visual tracking algorithms have been proposed in the last 2-3 decades, it remains a challenging problem for effective tracking with fast motion, deformation, occlusion et al. Under complex tracking conditions, most tracking models are not discriminative and adaptive enough. When the combined feature vectors are inputted to the visual models, this may lead to redundancy caused low efficiency and ambiguity caused poor performance. In this paper, an effective tracking algorithm is proposed to decontaminate features for each video sequence adaptively, where the visual modeling is treated as an optimization problem from the perspective of evolution. Every feature vector is compared to a biological individual and then decontaminated via classical evolutionary algorithms. With the optimized subsets of features, “Curse of Dimensionality” has been avoided whilst the accuracy of the visual model has been improved. The proposed algorithm has been tested on several publicly available datasets with various tracking challenges and benchmarked with a number of state-of-the-art approaches. The comprehensive experiments have demonstrated the efficacy of the proposed methodology.",
keywords = "visual tracking, feature subset, decontaminate, evolutionary algorithm",
author = "Qiaoyuan Liu and Yuru Wang and Minghao Yin and Jinchang Ren and Ruizhi Li",
note = "Copyright 2017 Society of Photo-Optical Instrumentation Engineers. One print or electronic copy may be made for personal use only. Systematic reproduction and distribution, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited.",
year = "2017",
month = "12",
day = "12",
doi = "10.1117/1.JEI.26.6.063025",
language = "English",
volume = "26",
journal = "Journal of Electronic Imaging",
issn = "1017-9909",
number = "6",

}

Decontaminate feature for tracking : adaptive tracking via evolutionary feature subset. / Liu, Qiaoyuan; Wang, Yuru; Yin, Minghao; Ren, Jinchang; Li, Ruizhi.

In: Journal of Electronic Imaging, Vol. 26, No. 6, 063025, 12.12.2017.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Decontaminate feature for tracking

T2 - Journal of Electronic Imaging

AU - Liu, Qiaoyuan

AU - Wang, Yuru

AU - Yin, Minghao

AU - Ren, Jinchang

AU - Li, Ruizhi

N1 - Copyright 2017 Society of Photo-Optical Instrumentation Engineers. One print or electronic copy may be made for personal use only. Systematic reproduction and distribution, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited.

PY - 2017/12/12

Y1 - 2017/12/12

N2 - Although various visual tracking algorithms have been proposed in the last 2-3 decades, it remains a challenging problem for effective tracking with fast motion, deformation, occlusion et al. Under complex tracking conditions, most tracking models are not discriminative and adaptive enough. When the combined feature vectors are inputted to the visual models, this may lead to redundancy caused low efficiency and ambiguity caused poor performance. In this paper, an effective tracking algorithm is proposed to decontaminate features for each video sequence adaptively, where the visual modeling is treated as an optimization problem from the perspective of evolution. Every feature vector is compared to a biological individual and then decontaminated via classical evolutionary algorithms. With the optimized subsets of features, “Curse of Dimensionality” has been avoided whilst the accuracy of the visual model has been improved. The proposed algorithm has been tested on several publicly available datasets with various tracking challenges and benchmarked with a number of state-of-the-art approaches. The comprehensive experiments have demonstrated the efficacy of the proposed methodology.

AB - Although various visual tracking algorithms have been proposed in the last 2-3 decades, it remains a challenging problem for effective tracking with fast motion, deformation, occlusion et al. Under complex tracking conditions, most tracking models are not discriminative and adaptive enough. When the combined feature vectors are inputted to the visual models, this may lead to redundancy caused low efficiency and ambiguity caused poor performance. In this paper, an effective tracking algorithm is proposed to decontaminate features for each video sequence adaptively, where the visual modeling is treated as an optimization problem from the perspective of evolution. Every feature vector is compared to a biological individual and then decontaminated via classical evolutionary algorithms. With the optimized subsets of features, “Curse of Dimensionality” has been avoided whilst the accuracy of the visual model has been improved. The proposed algorithm has been tested on several publicly available datasets with various tracking challenges and benchmarked with a number of state-of-the-art approaches. The comprehensive experiments have demonstrated the efficacy of the proposed methodology.

KW - visual tracking

KW - feature subset

KW - decontaminate

KW - evolutionary algorithm

UR - https://www.spiedigitallibrary.org/journals/journal-of-electronic-imaging

U2 - 10.1117/1.JEI.26.6.063025

DO - 10.1117/1.JEI.26.6.063025

M3 - Article

VL - 26

JO - Journal of Electronic Imaging

JF - Journal of Electronic Imaging

SN - 1017-9909

IS - 6

M1 - 063025

ER -