Constrained Superpixel Tracking

Lijun Wang, Huchuan Lu, Ming Hsuan Yang

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

In this paper, we propose a constrained graph labeling algorithm for visual tracking where nodes denote superpixels and edges encode the underlying spatial, temporal, and appearance fitness constraints. First, the spatial smoothness constraint, based on a transductive learning method, is enforced to leverage the latent manifold structure in feature space by investigating unlabeled superpixels in the current frame. Second, the appearance fitness constraint, which measures the probability of a superpixel being contained in the target region, is developed to incrementally induce a long-term appearance model. Third, the temporal smoothness constraint is proposed to construct a short-term appearance model of the target, which handles the drastic appearance change between consecutive frames. All these three constraints are incorporated in the proposed graph labeling algorithm such that induction and transduction, short- A nd long-term appearance models are combined, respectively. The foreground regions inferred by the proposed graph labeling method are used to guide the tracking process. Tracking results, in turn, facilitate more accurate online update by filtering out potential contaminated training samples. Both quantitative and qualitative evaluations on challenging tracking data sets show that the proposed constrained tracking algorithm performs favorably against the state-of-the-art methods.

Original languageEnglish
Article number7875137
Pages (from-to)1030-1041
Number of pages12
JournalIEEE Transactions on Cybernetics
Volume48
Issue number3
DOIs
Publication statusPublished - 2018 Mar

Fingerprint

Labeling

All Science Journal Classification (ASJC) codes

  • Software
  • Control and Systems Engineering
  • Information Systems
  • Human-Computer Interaction
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Wang, Lijun ; Lu, Huchuan ; Yang, Ming Hsuan. / Constrained Superpixel Tracking. In: IEEE Transactions on Cybernetics. 2018 ; Vol. 48, No. 3. pp. 1030-1041.
@article{cc0ecc4cf2ad47f095319140ca0a3cb2,
title = "Constrained Superpixel Tracking",
abstract = "In this paper, we propose a constrained graph labeling algorithm for visual tracking where nodes denote superpixels and edges encode the underlying spatial, temporal, and appearance fitness constraints. First, the spatial smoothness constraint, based on a transductive learning method, is enforced to leverage the latent manifold structure in feature space by investigating unlabeled superpixels in the current frame. Second, the appearance fitness constraint, which measures the probability of a superpixel being contained in the target region, is developed to incrementally induce a long-term appearance model. Third, the temporal smoothness constraint is proposed to construct a short-term appearance model of the target, which handles the drastic appearance change between consecutive frames. All these three constraints are incorporated in the proposed graph labeling algorithm such that induction and transduction, short- A nd long-term appearance models are combined, respectively. The foreground regions inferred by the proposed graph labeling method are used to guide the tracking process. Tracking results, in turn, facilitate more accurate online update by filtering out potential contaminated training samples. Both quantitative and qualitative evaluations on challenging tracking data sets show that the proposed constrained tracking algorithm performs favorably against the state-of-the-art methods.",
author = "Lijun Wang and Huchuan Lu and Yang, {Ming Hsuan}",
year = "2018",
month = "3",
doi = "10.1109/TCYB.2017.2675910",
language = "English",
volume = "48",
pages = "1030--1041",
journal = "IEEE Transactions on Cybernetics",
issn = "2168-2267",
publisher = "IEEE Advancing Technology for Humanity",
number = "3",

}

Constrained Superpixel Tracking. / Wang, Lijun; Lu, Huchuan; Yang, Ming Hsuan.

In: IEEE Transactions on Cybernetics, Vol. 48, No. 3, 7875137, 03.2018, p. 1030-1041.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Constrained Superpixel Tracking

AU - Wang, Lijun

AU - Lu, Huchuan

AU - Yang, Ming Hsuan

PY - 2018/3

Y1 - 2018/3

N2 - In this paper, we propose a constrained graph labeling algorithm for visual tracking where nodes denote superpixels and edges encode the underlying spatial, temporal, and appearance fitness constraints. First, the spatial smoothness constraint, based on a transductive learning method, is enforced to leverage the latent manifold structure in feature space by investigating unlabeled superpixels in the current frame. Second, the appearance fitness constraint, which measures the probability of a superpixel being contained in the target region, is developed to incrementally induce a long-term appearance model. Third, the temporal smoothness constraint is proposed to construct a short-term appearance model of the target, which handles the drastic appearance change between consecutive frames. All these three constraints are incorporated in the proposed graph labeling algorithm such that induction and transduction, short- A nd long-term appearance models are combined, respectively. The foreground regions inferred by the proposed graph labeling method are used to guide the tracking process. Tracking results, in turn, facilitate more accurate online update by filtering out potential contaminated training samples. Both quantitative and qualitative evaluations on challenging tracking data sets show that the proposed constrained tracking algorithm performs favorably against the state-of-the-art methods.

AB - In this paper, we propose a constrained graph labeling algorithm for visual tracking where nodes denote superpixels and edges encode the underlying spatial, temporal, and appearance fitness constraints. First, the spatial smoothness constraint, based on a transductive learning method, is enforced to leverage the latent manifold structure in feature space by investigating unlabeled superpixels in the current frame. Second, the appearance fitness constraint, which measures the probability of a superpixel being contained in the target region, is developed to incrementally induce a long-term appearance model. Third, the temporal smoothness constraint is proposed to construct a short-term appearance model of the target, which handles the drastic appearance change between consecutive frames. All these three constraints are incorporated in the proposed graph labeling algorithm such that induction and transduction, short- A nd long-term appearance models are combined, respectively. The foreground regions inferred by the proposed graph labeling method are used to guide the tracking process. Tracking results, in turn, facilitate more accurate online update by filtering out potential contaminated training samples. Both quantitative and qualitative evaluations on challenging tracking data sets show that the proposed constrained tracking algorithm performs favorably against the state-of-the-art methods.

UR - http://www.scopus.com/inward/record.url?scp=85016485217&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016485217&partnerID=8YFLogxK

U2 - 10.1109/TCYB.2017.2675910

DO - 10.1109/TCYB.2017.2675910

M3 - Article

AN - SCOPUS:85016485217

VL - 48

SP - 1030

EP - 1041

JO - IEEE Transactions on Cybernetics

JF - IEEE Transactions on Cybernetics

SN - 2168-2267

IS - 3

M1 - 7875137

ER -