A Robust Visual Tracker with a Coupled-Classifier Based on Multiple Representative Appearance Models

Deqian FU; Seong Tae JHANG

doi:10.1587/transinf.E96.D.1826

A Robust Visual Tracker with a Coupled-Classifier Based on Multiple Representative Appearance Models

Deqian FU, Seong Tae JHANG

Full Text Views

0

Cite this

Summary :

Aiming to alleviate the visual tracking problem of drift which reduces the abilities of almost all online visual trackers, a robust visual tracker (called CCMM tracker) is proposed with a coupled-classifier based on multiple representative appearance models. The coupled-classifier consists of root and head classifiers based on local sparse representation. The two classifiers collaborate to fulfil a tracking task within the Bayesian-based tracking framework, also to update their templates with a novel mechanism which tries to guarantee an update operation along the “right” orientation. Consequently, the tracker is more powerful in anti-interference. Meanwhile the multiple representative appearance models maintain features of the different submanifolds of the target appearance, which the target exhibited previously. The multiple models cooperatively support the coupled-classifier to recognize the target in challenging cases (such as persistent disturbance, vast change of appearance, and recovery from occlusion) with an effective strategy. The novel tracker proposed in this paper, by explicit inference, can reduce drift and handle frequent and drastic appearance variation of the target with cluttered background, which is demonstrated by the extensive experiments.

Publication: IEICE TRANSACTIONS on Information Vol.E96-D No.8 pp.1826-1835

Publication Date: 2013/08/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.E96.D.1826

Type of Manuscript: PAPER

Category: Image Recognition, Computer Vision

Authors

Deqian FU
Linyi University,The University of Suwon
Seong Tae JHANG
The University of Suwon

Keyword

coupled-classifier, representative appearance model, multiple appearance models, online visual tracking

Cite this

Copy

Deqian FU, Seong Tae JHANG, "A Robust Visual Tracker with a Coupled-Classifier Based on Multiple Representative Appearance Models" in IEICE TRANSACTIONS on Information, vol. E96-D, no. 8, pp. 1826-1835, August 2013, doi: 10.1587/transinf.E96.D.1826.
Abstract: Aiming to alleviate the visual tracking problem of drift which reduces the abilities of almost all online visual trackers, a robust visual tracker (called CCMM tracker) is proposed with a coupled-classifier based on multiple representative appearance models. The coupled-classifier consists of root and head classifiers based on local sparse representation. The two classifiers collaborate to fulfil a tracking task within the Bayesian-based tracking framework, also to update their templates with a novel mechanism which tries to guarantee an update operation along the “right” orientation. Consequently, the tracker is more powerful in anti-interference. Meanwhile the multiple representative appearance models maintain features of the different submanifolds of the target appearance, which the target exhibited previously. The multiple models cooperatively support the coupled-classifier to recognize the target in challenging cases (such as persistent disturbance, vast change of appearance, and recovery from occlusion) with an effective strategy. The novel tracker proposed in this paper, by explicit inference, can reduce drift and handle frequent and drastic appearance variation of the target with cluttered background, which is demonstrated by the extensive experiments.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E96.D.1826/_p

Copy

@ARTICLE{e96-d_8_1826,
author={Deqian FU, Seong Tae JHANG, },
journal={IEICE TRANSACTIONS on Information},
title={A Robust Visual Tracker with a Coupled-Classifier Based on Multiple Representative Appearance Models},
year={2013},
volume={E96-D},
number={8},
pages={1826-1835},
abstract={Aiming to alleviate the visual tracking problem of drift which reduces the abilities of almost all online visual trackers, a robust visual tracker (called CCMM tracker) is proposed with a coupled-classifier based on multiple representative appearance models. The coupled-classifier consists of root and head classifiers based on local sparse representation. The two classifiers collaborate to fulfil a tracking task within the Bayesian-based tracking framework, also to update their templates with a novel mechanism which tries to guarantee an update operation along the “right” orientation. Consequently, the tracker is more powerful in anti-interference. Meanwhile the multiple representative appearance models maintain features of the different submanifolds of the target appearance, which the target exhibited previously. The multiple models cooperatively support the coupled-classifier to recognize the target in challenging cases (such as persistent disturbance, vast change of appearance, and recovery from occlusion) with an effective strategy. The novel tracker proposed in this paper, by explicit inference, can reduce drift and handle frequent and drastic appearance variation of the target with cluttered background, which is demonstrated by the extensive experiments.},
keywords={},
doi={10.1587/transinf.E96.D.1826},
ISSN={1745-1361},
month={August},}

Copy

TY - JOUR
TI - A Robust Visual Tracker with a Coupled-Classifier Based on Multiple Representative Appearance Models
T2 - IEICE TRANSACTIONS on Information
SP - 1826
EP - 1835
AU - Deqian FU
AU - Seong Tae JHANG
PY - 2013
DO - 10.1587/transinf.E96.D.1826
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E96-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2013
AB - Aiming to alleviate the visual tracking problem of drift which reduces the abilities of almost all online visual trackers, a robust visual tracker (called CCMM tracker) is proposed with a coupled-classifier based on multiple representative appearance models. The coupled-classifier consists of root and head classifiers based on local sparse representation. The two classifiers collaborate to fulfil a tracking task within the Bayesian-based tracking framework, also to update their templates with a novel mechanism which tries to guarantee an update operation along the “right” orientation. Consequently, the tracker is more powerful in anti-interference. Meanwhile the multiple representative appearance models maintain features of the different submanifolds of the target appearance, which the target exhibited previously. The multiple models cooperatively support the coupled-classifier to recognize the target in challenging cases (such as persistent disturbance, vast change of appearance, and recovery from occlusion) with an effective strategy. The novel tracker proposed in this paper, by explicit inference, can reduce drift and handle frequent and drastic appearance variation of the target with cluttered background, which is demonstrated by the extensive experiments.
ER -