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...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| 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 |