Visual tracking based on transfer learning of deep salience information
In this paper, we propose a new visual tracking method in light of salience information and deep learning. Salience detection is used to exploit features with salient information of the image. Complicated representations of image features can be gained by the function of every layer in convolution n...
Guardado en:
Autores principales: | , , , |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
Institue of Optics and Electronics, Chinese Academy of Sciences
2020
|
Materias: | |
Acceso en línea: | https://doaj.org/article/e230fac26e2948d28ca973523f087436 |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:e230fac26e2948d28ca973523f087436 |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:e230fac26e2948d28ca973523f0874362021-11-10T10:02:56ZVisual tracking based on transfer learning of deep salience information2096-457910.29026/oea.2020.190018https://doaj.org/article/e230fac26e2948d28ca973523f0874362020-09-01T00:00:00Zhttp://www.oejournal.org/article/doi/10.29026/oea.2020.190018https://doaj.org/toc/2096-4579In this paper, we propose a new visual tracking method in light of salience information and deep learning. Salience detection is used to exploit features with salient information of the image. Complicated representations of image features can be gained by the function of every layer in convolution neural network (CNN). The characteristic of biology vision in attention-based salience is similar to the neuroscience features of convolution neural network. This motivates us to improve the representation ability of CNN with functions of salience detection. We adopt the fully-convolution networks (FCNs) to perform salience detection. We take parts of the network structure to perform salience extraction, which promotes the classification ability of the model. The network we propose shows great performance in tracking with the salient information. Compared with other excellent algorithms, our algorithm can track the target better in the open tracking datasets. We realize the 0.5592 accuracy on visual object tracking 2015 (VOT15) dataset. For unmanned aerial vehicle 123 (UAV123) dataset, the precision and success rate of our tracker is 0.710 and 0.429.Zuo HaoruiXu ZhiyongZhang JianlinJia GeInstitue of Optics and Electronics, Chinese Academy of Sciencesarticleconvolution neural networktransfer learningsalience detectionvisual trackingOptics. LightQC350-467ENOpto-Electronic Advances, Vol 3, Iss 9, Pp 190018-1-190018-11 (2020) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
convolution neural network transfer learning salience detection visual tracking Optics. Light QC350-467 |
spellingShingle |
convolution neural network transfer learning salience detection visual tracking Optics. Light QC350-467 Zuo Haorui Xu Zhiyong Zhang Jianlin Jia Ge Visual tracking based on transfer learning of deep salience information |
description |
In this paper, we propose a new visual tracking method in light of salience information and deep learning. Salience detection is used to exploit features with salient information of the image. Complicated representations of image features can be gained by the function of every layer in convolution neural network (CNN). The characteristic of biology vision in attention-based salience is similar to the neuroscience features of convolution neural network. This motivates us to improve the representation ability of CNN with functions of salience detection. We adopt the fully-convolution networks (FCNs) to perform salience detection. We take parts of the network structure to perform salience extraction, which promotes the classification ability of the model. The network we propose shows great performance in tracking with the salient information. Compared with other excellent algorithms, our algorithm can track the target better in the open tracking datasets. We realize the 0.5592 accuracy on visual object tracking 2015 (VOT15) dataset. For unmanned aerial vehicle 123 (UAV123) dataset, the precision and success rate of our tracker is 0.710 and 0.429. |
format |
article |
author |
Zuo Haorui Xu Zhiyong Zhang Jianlin Jia Ge |
author_facet |
Zuo Haorui Xu Zhiyong Zhang Jianlin Jia Ge |
author_sort |
Zuo Haorui |
title |
Visual tracking based on transfer learning of deep salience information |
title_short |
Visual tracking based on transfer learning of deep salience information |
title_full |
Visual tracking based on transfer learning of deep salience information |
title_fullStr |
Visual tracking based on transfer learning of deep salience information |
title_full_unstemmed |
Visual tracking based on transfer learning of deep salience information |
title_sort |
visual tracking based on transfer learning of deep salience information |
publisher |
Institue of Optics and Electronics, Chinese Academy of Sciences |
publishDate |
2020 |
url |
https://doaj.org/article/e230fac26e2948d28ca973523f087436 |
work_keys_str_mv |
AT zuohaorui visualtrackingbasedontransferlearningofdeepsalienceinformation AT xuzhiyong visualtrackingbasedontransferlearningofdeepsalienceinformation AT zhangjianlin visualtrackingbasedontransferlearningofdeepsalienceinformation AT jiage visualtrackingbasedontransferlearningofdeepsalienceinformation |
_version_ |
1718440149664137216 |