An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator
To prevent unmanned aerial vehicles (UAVs) from threatening public security, anti-UAV object tracking has become a critical issue in industrial and military applications. However, tracking UAV objects stably is still a challenging issue because the scenarios are complicated and the targets are gener...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-05-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/22/10/3701 |
_version_ | 1797495781045829632 |
---|---|
author | Feng Cheng Zhibo Liang Gaoliang Peng Shaohui Liu Sijue Li Mengyu Ji |
author_facet | Feng Cheng Zhibo Liang Gaoliang Peng Shaohui Liu Sijue Li Mengyu Ji |
author_sort | Feng Cheng |
collection | DOAJ |
description | To prevent unmanned aerial vehicles (UAVs) from threatening public security, anti-UAV object tracking has become a critical issue in industrial and military applications. However, tracking UAV objects stably is still a challenging issue because the scenarios are complicated and the targets are generally small. In this article, a novel long-term tracking architecture composed of a Siamese network and re-detection (SiamAD) is proposed to efficiently locate UAV targets in diverse surroundings. Specifically, a new hybrid attention mechanism module is exploited to conduct more discriminative feature representation and is incorporated into a Siamese network. At the same time, the attention-based Siamese network fuses multilevel features for accurately tracking the target. We further introduce a hierarchical discriminator for checking the reliability of targeting, and a discriminator-based redetection network is utilized for correcting tracking failures. To effectively catch up with the appearance changes of UAVs, a template updating strategy is developed in long-term tracking tasks. Our model surpasses many state-of-the-art models on the anti-UAV benchmark. In particular, the proposed method can achieve 13.7% and 16.5% improvements in success rate and precision rate, respectively, compared with the strong baseline SiamRPN++. |
first_indexed | 2024-03-10T01:54:33Z |
format | Article |
id | doaj.art-f909c377371b4140b7a1d012cbab2a85 |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-10T01:54:33Z |
publishDate | 2022-05-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-f909c377371b4140b7a1d012cbab2a852023-11-23T12:59:41ZengMDPI AGSensors1424-82202022-05-012210370110.3390/s22103701An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical DiscriminatorFeng Cheng0Zhibo Liang1Gaoliang Peng2Shaohui Liu3Sijue Li4Mengyu Ji5State Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150000, ChinaDepartment of Computer Science and Technology, Harbin Institute of Technology, Harbin 150000, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150000, ChinaDepartment of Computer Science and Technology, Harbin Institute of Technology, Harbin 150000, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150000, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150000, ChinaTo prevent unmanned aerial vehicles (UAVs) from threatening public security, anti-UAV object tracking has become a critical issue in industrial and military applications. However, tracking UAV objects stably is still a challenging issue because the scenarios are complicated and the targets are generally small. In this article, a novel long-term tracking architecture composed of a Siamese network and re-detection (SiamAD) is proposed to efficiently locate UAV targets in diverse surroundings. Specifically, a new hybrid attention mechanism module is exploited to conduct more discriminative feature representation and is incorporated into a Siamese network. At the same time, the attention-based Siamese network fuses multilevel features for accurately tracking the target. We further introduce a hierarchical discriminator for checking the reliability of targeting, and a discriminator-based redetection network is utilized for correcting tracking failures. To effectively catch up with the appearance changes of UAVs, a template updating strategy is developed in long-term tracking tasks. Our model surpasses many state-of-the-art models on the anti-UAV benchmark. In particular, the proposed method can achieve 13.7% and 16.5% improvements in success rate and precision rate, respectively, compared with the strong baseline SiamRPN++.https://www.mdpi.com/1424-8220/22/10/3701anti-UAVlong-term trackingattention mechanismdiscriminatorSiamese network |
spellingShingle | Feng Cheng Zhibo Liang Gaoliang Peng Shaohui Liu Sijue Li Mengyu Ji An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator Sensors anti-UAV long-term tracking attention mechanism discriminator Siamese network |
title | An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator |
title_full | An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator |
title_fullStr | An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator |
title_full_unstemmed | An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator |
title_short | An Anti-UAV Long-Term Tracking Method with Hybrid Attention Mechanism and Hierarchical Discriminator |
title_sort | anti uav long term tracking method with hybrid attention mechanism and hierarchical discriminator |
topic | anti-UAV long-term tracking attention mechanism discriminator Siamese network |
url | https://www.mdpi.com/1424-8220/22/10/3701 |
work_keys_str_mv | AT fengcheng anantiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT zhiboliang anantiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT gaoliangpeng anantiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT shaohuiliu anantiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT sijueli anantiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT mengyuji anantiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT fengcheng antiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT zhiboliang antiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT gaoliangpeng antiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT shaohuiliu antiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT sijueli antiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator AT mengyuji antiuavlongtermtrackingmethodwithhybridattentionmechanismandhierarchicaldiscriminator |