L0-regularized object representation for visual tracking

Jinshan Pan, Jongwoo Lim, Zhixun Su, Ming Hsuan Yang

Research output: Contribution to conferencePaper

12 Citations (Scopus)

Abstract

In this paper, we propose a robust visual tracking method by L0-regularized prior in a particle filter framework. In contrast to existing methods, the proposed method employs L0 norm to regularize the linear coefficients of incrementally updated linear basis. The sparsity constraint enables the tracker to effectively handle difficult challenges, such as occlusion or image corruption. To achieve realtime processing, we propose a fast and efficient numerical algorithm for solving the proposed L0-regularized model. Although it is an NP-hard problem, the proposed accelerated proximal gradient (APG) approach is guaranteed to converge to a solution quickly. Extensive experimental results on challenging video sequences demonstrate that the proposed method achieves state-of-the-art results both in accuracy and speed.

Original languageEnglish
Publication statusPublished - 2014 Jan 1
Event25th British Machine Vision Conference, BMVC 2014 - Nottingham, United Kingdom
Duration: 2014 Sep 12014 Sep 5

Conference

Conference25th British Machine Vision Conference, BMVC 2014
CountryUnited Kingdom
CityNottingham
Period14/9/114/9/5

Fingerprint

Computational complexity
Processing

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

Pan, J., Lim, J., Su, Z., & Yang, M. H. (2014). L0-regularized object representation for visual tracking. Paper presented at 25th British Machine Vision Conference, BMVC 2014, Nottingham, United Kingdom.
Pan, Jinshan ; Lim, Jongwoo ; Su, Zhixun ; Yang, Ming Hsuan. / L0-regularized object representation for visual tracking. Paper presented at 25th British Machine Vision Conference, BMVC 2014, Nottingham, United Kingdom.
@conference{1789228ec8764b6b8be10f150de9ddb4,
title = "L0-regularized object representation for visual tracking",
abstract = "In this paper, we propose a robust visual tracking method by L0-regularized prior in a particle filter framework. In contrast to existing methods, the proposed method employs L0 norm to regularize the linear coefficients of incrementally updated linear basis. The sparsity constraint enables the tracker to effectively handle difficult challenges, such as occlusion or image corruption. To achieve realtime processing, we propose a fast and efficient numerical algorithm for solving the proposed L0-regularized model. Although it is an NP-hard problem, the proposed accelerated proximal gradient (APG) approach is guaranteed to converge to a solution quickly. Extensive experimental results on challenging video sequences demonstrate that the proposed method achieves state-of-the-art results both in accuracy and speed.",
author = "Jinshan Pan and Jongwoo Lim and Zhixun Su and Yang, {Ming Hsuan}",
year = "2014",
month = "1",
day = "1",
language = "English",
note = "25th British Machine Vision Conference, BMVC 2014 ; Conference date: 01-09-2014 Through 05-09-2014",

}

Pan, J, Lim, J, Su, Z & Yang, MH 2014, 'L0-regularized object representation for visual tracking' Paper presented at 25th British Machine Vision Conference, BMVC 2014, Nottingham, United Kingdom, 14/9/1 - 14/9/5, .

L0-regularized object representation for visual tracking. / Pan, Jinshan; Lim, Jongwoo; Su, Zhixun; Yang, Ming Hsuan.

2014. Paper presented at 25th British Machine Vision Conference, BMVC 2014, Nottingham, United Kingdom.

Research output: Contribution to conferencePaper

TY - CONF

T1 - L0-regularized object representation for visual tracking

AU - Pan, Jinshan

AU - Lim, Jongwoo

AU - Su, Zhixun

AU - Yang, Ming Hsuan

PY - 2014/1/1

Y1 - 2014/1/1

N2 - In this paper, we propose a robust visual tracking method by L0-regularized prior in a particle filter framework. In contrast to existing methods, the proposed method employs L0 norm to regularize the linear coefficients of incrementally updated linear basis. The sparsity constraint enables the tracker to effectively handle difficult challenges, such as occlusion or image corruption. To achieve realtime processing, we propose a fast and efficient numerical algorithm for solving the proposed L0-regularized model. Although it is an NP-hard problem, the proposed accelerated proximal gradient (APG) approach is guaranteed to converge to a solution quickly. Extensive experimental results on challenging video sequences demonstrate that the proposed method achieves state-of-the-art results both in accuracy and speed.

AB - In this paper, we propose a robust visual tracking method by L0-regularized prior in a particle filter framework. In contrast to existing methods, the proposed method employs L0 norm to regularize the linear coefficients of incrementally updated linear basis. The sparsity constraint enables the tracker to effectively handle difficult challenges, such as occlusion or image corruption. To achieve realtime processing, we propose a fast and efficient numerical algorithm for solving the proposed L0-regularized model. Although it is an NP-hard problem, the proposed accelerated proximal gradient (APG) approach is guaranteed to converge to a solution quickly. Extensive experimental results on challenging video sequences demonstrate that the proposed method achieves state-of-the-art results both in accuracy and speed.

UR - http://www.scopus.com/inward/record.url?scp=84919741173&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84919741173&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84919741173

ER -

Pan J, Lim J, Su Z, Yang MH. L0-regularized object representation for visual tracking. 2014. Paper presented at 25th British Machine Vision Conference, BMVC 2014, Nottingham, United Kingdom.