A Target Tracking Method Based on a Pyramid Channel Attention Mechanism

To enhance the tracking performance of transformer-based trackers in complex scenes, we propose a novel visual object tracking method that incorporates three key components: a pyramid channel attention mechanism, a hierarchical cross-attention structure, and an attention-guided multi-layer perceptro...

Full description

Saved in:
Bibliographic Details
Main Authors: Dongxuan Zhao, Yunsong Li, Jiaxing Li, Xiping Duan, Ning Ma, Yuhe Wang
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/10/3214
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849327016565800960
author Dongxuan Zhao
Yunsong Li
Jiaxing Li
Xiping Duan
Ning Ma
Yuhe Wang
author_facet Dongxuan Zhao
Yunsong Li
Jiaxing Li
Xiping Duan
Ning Ma
Yuhe Wang
author_sort Dongxuan Zhao
collection DOAJ
description To enhance the tracking performance of transformer-based trackers in complex scenes, we propose a novel visual object tracking method that incorporates three key components: a pyramid channel attention mechanism, a hierarchical cross-attention structure, and an attention-guided multi-layer perceptron. The pyramid channel attention mechanism dynamically enhances informative feature channels across different scales, while the hierarchical cross-attention structure facilitates effective feature interaction. The attention-guided multi-layer perceptron introduces nonlinear transformations under attention guidance to improve feature representation. Experimental results on benchmark datasets demonstrate the superior performance of the proposed method.
format Article
id doaj-art-c02045de8d0b443d9739d884b7cbdf56
institution Kabale University
issn 1424-8220
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-c02045de8d0b443d9739d884b7cbdf562025-08-20T03:47:58ZengMDPI AGSensors1424-82202025-05-012510321410.3390/s25103214A Target Tracking Method Based on a Pyramid Channel Attention MechanismDongxuan Zhao0Yunsong Li1Jiaxing Li2Xiping Duan3Ning Ma4Yuhe Wang5School of Computer Science and Information Engineering, Harbin Normal University, No. 1 Shida Road, Limin Economic Development Zone, Harbin 150025, ChinaSchool of Computer Science and Information Engineering, Harbin Normal University, No. 1 Shida Road, Limin Economic Development Zone, Harbin 150025, ChinaSchool of Computer Science and Information Engineering, Harbin Normal University, No. 1 Shida Road, Limin Economic Development Zone, Harbin 150025, ChinaSchool of Computer Science and Information Engineering, Harbin Normal University, No. 1 Shida Road, Limin Economic Development Zone, Harbin 150025, ChinaSchool of Computer Science and Information Engineering, Harbin Normal University, No. 1 Shida Road, Limin Economic Development Zone, Harbin 150025, ChinaSchool of Computer Science and Information Engineering, Harbin Normal University, No. 1 Shida Road, Limin Economic Development Zone, Harbin 150025, ChinaTo enhance the tracking performance of transformer-based trackers in complex scenes, we propose a novel visual object tracking method that incorporates three key components: a pyramid channel attention mechanism, a hierarchical cross-attention structure, and an attention-guided multi-layer perceptron. The pyramid channel attention mechanism dynamically enhances informative feature channels across different scales, while the hierarchical cross-attention structure facilitates effective feature interaction. The attention-guided multi-layer perceptron introduces nonlinear transformations under attention guidance to improve feature representation. Experimental results on benchmark datasets demonstrate the superior performance of the proposed method.https://www.mdpi.com/1424-8220/25/10/3214target trackingdeep learningtransformerfeature extractionfeature fusion
spellingShingle Dongxuan Zhao
Yunsong Li
Jiaxing Li
Xiping Duan
Ning Ma
Yuhe Wang
A Target Tracking Method Based on a Pyramid Channel Attention Mechanism
Sensors
target tracking
deep learning
transformer
feature extraction
feature fusion
title A Target Tracking Method Based on a Pyramid Channel Attention Mechanism
title_full A Target Tracking Method Based on a Pyramid Channel Attention Mechanism
title_fullStr A Target Tracking Method Based on a Pyramid Channel Attention Mechanism
title_full_unstemmed A Target Tracking Method Based on a Pyramid Channel Attention Mechanism
title_short A Target Tracking Method Based on a Pyramid Channel Attention Mechanism
title_sort target tracking method based on a pyramid channel attention mechanism
topic target tracking
deep learning
transformer
feature extraction
feature fusion
url https://www.mdpi.com/1424-8220/25/10/3214
work_keys_str_mv AT dongxuanzhao atargettrackingmethodbasedonapyramidchannelattentionmechanism
AT yunsongli atargettrackingmethodbasedonapyramidchannelattentionmechanism
AT jiaxingli atargettrackingmethodbasedonapyramidchannelattentionmechanism
AT xipingduan atargettrackingmethodbasedonapyramidchannelattentionmechanism
AT ningma atargettrackingmethodbasedonapyramidchannelattentionmechanism
AT yuhewang atargettrackingmethodbasedonapyramidchannelattentionmechanism
AT dongxuanzhao targettrackingmethodbasedonapyramidchannelattentionmechanism
AT yunsongli targettrackingmethodbasedonapyramidchannelattentionmechanism
AT jiaxingli targettrackingmethodbasedonapyramidchannelattentionmechanism
AT xipingduan targettrackingmethodbasedonapyramidchannelattentionmechanism
AT ningma targettrackingmethodbasedonapyramidchannelattentionmechanism
AT yuhewang targettrackingmethodbasedonapyramidchannelattentionmechanism