Hyperspectral Attention Network for Object Tracking
Hyperspectral video provides rich spatial and spectral information, which is crucial for object tracking in complex scenarios. Despite extensive research, existing methods often face an inherent trade-off between rich spectral information and redundant noisy information. This dilemma arises from the...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-09-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/19/6178 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850284396757647360 |
|---|---|
| author | Shuangjiang Yu Jianjun Ni Shuai Fu Tao Qu |
| author_facet | Shuangjiang Yu Jianjun Ni Shuai Fu Tao Qu |
| author_sort | Shuangjiang Yu |
| collection | DOAJ |
| description | Hyperspectral video provides rich spatial and spectral information, which is crucial for object tracking in complex scenarios. Despite extensive research, existing methods often face an inherent trade-off between rich spectral information and redundant noisy information. This dilemma arises from the efficient utilization of hyperspectral image data channels. To alleviate this problem, this paper introduces a hierarchical spectral attention network for hyperspectral object tracking. We employ a spectral band attention mechanism with adaptive soft threshold to examine the correlations across spectral bands, which integrates the information available in various spectral bands and eliminates redundant information. Moreover, we integrate spectral attention into a hierarchical tracking network to improve the integration of spectral and spatial information. The experimental results on entire public hyperspectral competition dataset WHISPER2020 show the superior performance of our proposed method compared with that of several related methods in visual effects and objective evaluation. |
| format | Article |
| id | doaj-art-9e4595fe6ccc47edbf8ef3f52474eb39 |
| institution | OA Journals |
| issn | 1424-8220 |
| language | English |
| publishDate | 2024-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-9e4595fe6ccc47edbf8ef3f52474eb392025-08-20T01:47:34ZengMDPI AGSensors1424-82202024-09-012419617810.3390/s24196178Hyperspectral Attention Network for Object TrackingShuangjiang Yu0Jianjun Ni1Shuai Fu2Tao Qu3Beijing Institute of Space Mechanics and Electricity, Beijing 100094, ChinaBeijing Institute of Space Mechanics and Electricity, Beijing 100094, ChinaBeijing Institute of Space Mechanics and Electricity, Beijing 100094, ChinaSchool of Computer Science, Wuhan University, Wuhan 430072, ChinaHyperspectral video provides rich spatial and spectral information, which is crucial for object tracking in complex scenarios. Despite extensive research, existing methods often face an inherent trade-off between rich spectral information and redundant noisy information. This dilemma arises from the efficient utilization of hyperspectral image data channels. To alleviate this problem, this paper introduces a hierarchical spectral attention network for hyperspectral object tracking. We employ a spectral band attention mechanism with adaptive soft threshold to examine the correlations across spectral bands, which integrates the information available in various spectral bands and eliminates redundant information. Moreover, we integrate spectral attention into a hierarchical tracking network to improve the integration of spectral and spatial information. The experimental results on entire public hyperspectral competition dataset WHISPER2020 show the superior performance of our proposed method compared with that of several related methods in visual effects and objective evaluation.https://www.mdpi.com/1424-8220/24/19/6178hyperspectral videosobject trackingmultiscale featuresspectral attention |
| spellingShingle | Shuangjiang Yu Jianjun Ni Shuai Fu Tao Qu Hyperspectral Attention Network for Object Tracking Sensors hyperspectral videos object tracking multiscale features spectral attention |
| title | Hyperspectral Attention Network for Object Tracking |
| title_full | Hyperspectral Attention Network for Object Tracking |
| title_fullStr | Hyperspectral Attention Network for Object Tracking |
| title_full_unstemmed | Hyperspectral Attention Network for Object Tracking |
| title_short | Hyperspectral Attention Network for Object Tracking |
| title_sort | hyperspectral attention network for object tracking |
| topic | hyperspectral videos object tracking multiscale features spectral attention |
| url | https://www.mdpi.com/1424-8220/24/19/6178 |
| work_keys_str_mv | AT shuangjiangyu hyperspectralattentionnetworkforobjecttracking AT jianjunni hyperspectralattentionnetworkforobjecttracking AT shuaifu hyperspectralattentionnetworkforobjecttracking AT taoqu hyperspectralattentionnetworkforobjecttracking |