Cargando…

Visual Tracking Based on Extreme Learning Machine and Sparse Representation

The existing sparse representation-based visual trackers mostly suffer from both being time consuming and having poor robustness problems. To address these issues, a novel tracking method is presented via combining sparse representation and an emerging learning technique, namely extreme learning mac...

Descripción completa

Detalles Bibliográficos
Autores principales: Wang, Baoxian, Tang, Linbo, Yang, Jinglin, Zhao, Baojun, Wang, Shuigen
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2015
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4634458/
https://www.ncbi.nlm.nih.gov/pubmed/26506359
http://dx.doi.org/10.3390/s151026877
_version_ 1782399360706805760
author Wang, Baoxian
Tang, Linbo
Yang, Jinglin
Zhao, Baojun
Wang, Shuigen
author_facet Wang, Baoxian
Tang, Linbo
Yang, Jinglin
Zhao, Baojun
Wang, Shuigen
author_sort Wang, Baoxian
collection PubMed
description The existing sparse representation-based visual trackers mostly suffer from both being time consuming and having poor robustness problems. To address these issues, a novel tracking method is presented via combining sparse representation and an emerging learning technique, namely extreme learning machine (ELM). Specifically, visual tracking can be divided into two consecutive processes. Firstly, ELM is utilized to find the optimal separate hyperplane between the target observations and background ones. Thus, the trained ELM classification function is able to remove most of the candidate samples related to background contents efficiently, thereby reducing the total computational cost of the following sparse representation. Secondly, to further combine ELM and sparse representation, the resultant confidence values (i.e., probabilities to be a target) of samples on the ELM classification function are used to construct a new manifold learning constraint term of the sparse representation framework, which tends to achieve robuster results. Moreover, the accelerated proximal gradient method is used for deriving the optimal solution (in matrix form) of the constrained sparse tracking model. Additionally, the matrix form solution allows the candidate samples to be calculated in parallel, thereby leading to a higher efficiency. Experiments demonstrate the effectiveness of the proposed tracker.
format Online
Article
Text
id pubmed-4634458
institution National Center for Biotechnology Information
language English
publishDate 2015
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-46344582015-11-23 Visual Tracking Based on Extreme Learning Machine and Sparse Representation Wang, Baoxian Tang, Linbo Yang, Jinglin Zhao, Baojun Wang, Shuigen Sensors (Basel) Article The existing sparse representation-based visual trackers mostly suffer from both being time consuming and having poor robustness problems. To address these issues, a novel tracking method is presented via combining sparse representation and an emerging learning technique, namely extreme learning machine (ELM). Specifically, visual tracking can be divided into two consecutive processes. Firstly, ELM is utilized to find the optimal separate hyperplane between the target observations and background ones. Thus, the trained ELM classification function is able to remove most of the candidate samples related to background contents efficiently, thereby reducing the total computational cost of the following sparse representation. Secondly, to further combine ELM and sparse representation, the resultant confidence values (i.e., probabilities to be a target) of samples on the ELM classification function are used to construct a new manifold learning constraint term of the sparse representation framework, which tends to achieve robuster results. Moreover, the accelerated proximal gradient method is used for deriving the optimal solution (in matrix form) of the constrained sparse tracking model. Additionally, the matrix form solution allows the candidate samples to be calculated in parallel, thereby leading to a higher efficiency. Experiments demonstrate the effectiveness of the proposed tracker. MDPI 2015-10-22 /pmc/articles/PMC4634458/ /pubmed/26506359 http://dx.doi.org/10.3390/s151026877 Text en © 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Wang, Baoxian
Tang, Linbo
Yang, Jinglin
Zhao, Baojun
Wang, Shuigen
Visual Tracking Based on Extreme Learning Machine and Sparse Representation
title Visual Tracking Based on Extreme Learning Machine and Sparse Representation
title_full Visual Tracking Based on Extreme Learning Machine and Sparse Representation
title_fullStr Visual Tracking Based on Extreme Learning Machine and Sparse Representation
title_full_unstemmed Visual Tracking Based on Extreme Learning Machine and Sparse Representation
title_short Visual Tracking Based on Extreme Learning Machine and Sparse Representation
title_sort visual tracking based on extreme learning machine and sparse representation
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4634458/
https://www.ncbi.nlm.nih.gov/pubmed/26506359
http://dx.doi.org/10.3390/s151026877
work_keys_str_mv AT wangbaoxian visualtrackingbasedonextremelearningmachineandsparserepresentation
AT tanglinbo visualtrackingbasedonextremelearningmachineandsparserepresentation
AT yangjinglin visualtrackingbasedonextremelearningmachineandsparserepresentation
AT zhaobaojun visualtrackingbasedonextremelearningmachineandsparserepresentation
AT wangshuigen visualtrackingbasedonextremelearningmachineandsparserepresentation