A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes
In recent years, target tracking algorithms based on deep learning have realized significant progress, especially the Siamese neural network structure, which has a simple structure and excellent scalability. Although these methods provide excellent generalization capabilities, they fail to perform t...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-09-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/11/19/3125 |
_version_ | 1797479828195115008 |
---|---|
author | Yu Wang Zhutian Yang Wei Yang Jiamin Yang |
author_facet | Yu Wang Zhutian Yang Wei Yang Jiamin Yang |
author_sort | Yu Wang |
collection | DOAJ |
description | In recent years, target tracking algorithms based on deep learning have realized significant progress, especially the Siamese neural network structure, which has a simple structure and excellent scalability. Although these methods provide excellent generalization capabilities, they fail to perform the task of learning target information discrimination smoothly due to being affected by distractors such as background clutter, occlusion, and target size. To solve this problem, in this paper we propose a newly improved Siamese network target tracking algorithm based on an attention mechanism. We introduce a channel attention module and a spatial attention module into the original network to improve the problem of insufficient semantic extraction ability of the convolutional layer of the tracking algorithm in complex environments. A channel attention mechanism enhances the feature extraction ability by using the network to learn the importance of each channel and establish the relationship between channels, while a spatial attention mechanism strengthens the feature extraction ability by establishing the importance of spatial position in locating the target or carrying out a certain degree of deformation. In this paper, the above two models are combined to improve the robustness of trackers without sacrificing tracking speed. We conduct a comprehensive experiment on the Object Tracking Benchmark dataset. The experimental results show that our algorithm outperforms other real-time trackers in both accuracy and robustness in most complex environments. |
first_indexed | 2024-03-09T21:51:25Z |
format | Article |
id | doaj.art-f403ad3a59234df99d56f8a117b0d8a3 |
institution | Directory Open Access Journal |
issn | 2079-9292 |
language | English |
last_indexed | 2024-03-09T21:51:25Z |
publishDate | 2022-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Electronics |
spelling | doaj.art-f403ad3a59234df99d56f8a117b0d8a32023-11-23T20:06:42ZengMDPI AGElectronics2079-92922022-09-011119312510.3390/electronics11193125A Novel Target Tracking Scheme Based on Attention Mechanism in Complex ScenesYu Wang0Zhutian Yang1Wei Yang2Jiamin Yang3School of Electronics and Information Engineering, Harbin Institute of Technology, Harbin 150080, ChinaSchool of Electronics and Information Engineering, Harbin Institute of Technology, Harbin 150080, ChinaNanjing Electronic Equipment Institute, Nanjing 210013, ChinaSchool of Information Science and Engineering, Southeast University, Nanjing 211189, ChinaIn recent years, target tracking algorithms based on deep learning have realized significant progress, especially the Siamese neural network structure, which has a simple structure and excellent scalability. Although these methods provide excellent generalization capabilities, they fail to perform the task of learning target information discrimination smoothly due to being affected by distractors such as background clutter, occlusion, and target size. To solve this problem, in this paper we propose a newly improved Siamese network target tracking algorithm based on an attention mechanism. We introduce a channel attention module and a spatial attention module into the original network to improve the problem of insufficient semantic extraction ability of the convolutional layer of the tracking algorithm in complex environments. A channel attention mechanism enhances the feature extraction ability by using the network to learn the importance of each channel and establish the relationship between channels, while a spatial attention mechanism strengthens the feature extraction ability by establishing the importance of spatial position in locating the target or carrying out a certain degree of deformation. In this paper, the above two models are combined to improve the robustness of trackers without sacrificing tracking speed. We conduct a comprehensive experiment on the Object Tracking Benchmark dataset. The experimental results show that our algorithm outperforms other real-time trackers in both accuracy and robustness in most complex environments.https://www.mdpi.com/2079-9292/11/19/3125visual object trackingSiamese networkdeep learningfull convolutional neural networkattentional mechanism |
spellingShingle | Yu Wang Zhutian Yang Wei Yang Jiamin Yang A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes Electronics visual object tracking Siamese network deep learning full convolutional neural network attentional mechanism |
title | A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes |
title_full | A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes |
title_fullStr | A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes |
title_full_unstemmed | A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes |
title_short | A Novel Target Tracking Scheme Based on Attention Mechanism in Complex Scenes |
title_sort | novel target tracking scheme based on attention mechanism in complex scenes |
topic | visual object tracking Siamese network deep learning full convolutional neural network attentional mechanism |
url | https://www.mdpi.com/2079-9292/11/19/3125 |
work_keys_str_mv | AT yuwang anoveltargettrackingschemebasedonattentionmechanismincomplexscenes AT zhutianyang anoveltargettrackingschemebasedonattentionmechanismincomplexscenes AT weiyang anoveltargettrackingschemebasedonattentionmechanismincomplexscenes AT jiaminyang anoveltargettrackingschemebasedonattentionmechanismincomplexscenes AT yuwang noveltargettrackingschemebasedonattentionmechanismincomplexscenes AT zhutianyang noveltargettrackingschemebasedonattentionmechanismincomplexscenes AT weiyang noveltargettrackingschemebasedonattentionmechanismincomplexscenes AT jiaminyang noveltargettrackingschemebasedonattentionmechanismincomplexscenes |