Aware Distribute and Sparse Network for Infrared Small Target Detection
Deep learning has achieved tremendous success in the field of object detection. The efficient detection of infrared small targets using deep learning methods remains a challenging task. Infrared small targets are often detected in high-resolution features. Extracting high-level semantic features lay...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10459168/ |
| _version_ | 1827308240207085568 |
|---|---|
| author | Yansong Song Boxiao Wang Keyan Dong |
| author_facet | Yansong Song Boxiao Wang Keyan Dong |
| author_sort | Yansong Song |
| collection | DOAJ |
| description | Deep learning has achieved tremendous success in the field of object detection. The efficient detection of infrared small targets using deep learning methods remains a challenging task. Infrared small targets are often detected in high-resolution features. Extracting high-level semantic features layer by layer in the network may lead to the loss of deep-layer targets. However, performing global detection on high-resolution feature maps results in high computational costs. To address this issue, we propose the aware distribute and sparse network (ADSNet) to preserve deep-layer small target features while accelerating inference speed. Specifically, we design the aware fusion distribute module (AFD) to aggregate global features and enhance the representation capability of deep-layer features. Subsequently, the aware cascaded sparse module (ACS) is utilized to guide step-by-step high-resolution feature sparsification. Experimental results demonstrate that the proposed method achieves accurate segmentation in various detection scenarios and for diverse target morphologies, effectively suppressing false alarms while controlling computational expenses. Ablation experiments further validate the effectiveness of each component. |
| first_indexed | 2024-04-24T18:55:01Z |
| format | Article |
| id | doaj.art-553235339f2b497d86ad98fab9c26ba7 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-24T18:55:01Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-553235339f2b497d86ad98fab9c26ba72024-03-26T17:43:54ZengIEEEIEEE Access2169-35362024-01-0112405344054310.1109/ACCESS.2024.337343610459168Aware Distribute and Sparse Network for Infrared Small Target DetectionYansong Song0Boxiao Wang1https://orcid.org/0009-0005-2577-8710Keyan Dong2School of Electro-Optical Engineering, Changchun University of Science and Technology, Changchun, ChinaSchool of Electro-Optical Engineering, Changchun University of Science and Technology, Changchun, ChinaSchool of Electro-Optical Engineering, Changchun University of Science and Technology, Changchun, ChinaDeep learning has achieved tremendous success in the field of object detection. The efficient detection of infrared small targets using deep learning methods remains a challenging task. Infrared small targets are often detected in high-resolution features. Extracting high-level semantic features layer by layer in the network may lead to the loss of deep-layer targets. However, performing global detection on high-resolution feature maps results in high computational costs. To address this issue, we propose the aware distribute and sparse network (ADSNet) to preserve deep-layer small target features while accelerating inference speed. Specifically, we design the aware fusion distribute module (AFD) to aggregate global features and enhance the representation capability of deep-layer features. Subsequently, the aware cascaded sparse module (ACS) is utilized to guide step-by-step high-resolution feature sparsification. Experimental results demonstrate that the proposed method achieves accurate segmentation in various detection scenarios and for diverse target morphologies, effectively suppressing false alarms while controlling computational expenses. Ablation experiments further validate the effectiveness of each component.https://ieeexplore.ieee.org/document/10459168/Object detectioninfrared imaginginfrared small target detectionfeature fusion |
| spellingShingle | Yansong Song Boxiao Wang Keyan Dong Aware Distribute and Sparse Network for Infrared Small Target Detection IEEE Access Object detection infrared imaging infrared small target detection feature fusion |
| title | Aware Distribute and Sparse Network for Infrared Small Target Detection |
| title_full | Aware Distribute and Sparse Network for Infrared Small Target Detection |
| title_fullStr | Aware Distribute and Sparse Network for Infrared Small Target Detection |
| title_full_unstemmed | Aware Distribute and Sparse Network for Infrared Small Target Detection |
| title_short | Aware Distribute and Sparse Network for Infrared Small Target Detection |
| title_sort | aware distribute and sparse network for infrared small target detection |
| topic | Object detection infrared imaging infrared small target detection feature fusion |
| url | https://ieeexplore.ieee.org/document/10459168/ |
| work_keys_str_mv | AT yansongsong awaredistributeandsparsenetworkforinfraredsmalltargetdetection AT boxiaowang awaredistributeandsparsenetworkforinfraredsmalltargetdetection AT keyandong awaredistributeandsparsenetworkforinfraredsmalltargetdetection |