Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2505.21370 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912398235926528 |
|---|---|
| author | Wang, Xinyuan Peng, Lian Li, Xiangcheng He, Yilin U, KinTak |
| author_facet | Wang, Xinyuan Peng, Lian Li, Xiangcheng He, Yilin U, KinTak |
| contents | Object detection in remote sensing imagery remains a challenging task due to extreme scale variation, dense object distributions, and cluttered backgrounds. While recent detectors such as YOLOv8 have shown promising results, their backbone architectures lack explicit mechanisms to guide multi-scale feature refinement, limiting performance on high-resolution aerial data. In this work, we propose YOLO-SPCI, an attention-enhanced detection framework that introduces a lightweight Selective-Perspective-Class Integration (SPCI) module to improve feature representation. The SPCI module integrates three components: a Selective Stream Gate (SSG) for adaptive regulation of global feature flow, a Perspective Fusion Module (PFM) for context-aware multi-scale integration, and a Class Discrimination Module (CDM) to enhance inter-class separability. We embed two SPCI blocks into the P3 and P5 stages of the YOLOv8 backbone, enabling effective refinement while preserving compatibility with the original neck and head. Experiments on the NWPU VHR-10 dataset demonstrate that YOLO-SPCI achieves superior performance compared to state-of-the-art detectors. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_21370 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | YOLO-SPCI: Enhancing Remote Sensing Object Detection via Selective-Perspective-Class Integration Wang, Xinyuan Peng, Lian Li, Xiangcheng He, Yilin U, KinTak Computer Vision and Pattern Recognition Object detection in remote sensing imagery remains a challenging task due to extreme scale variation, dense object distributions, and cluttered backgrounds. While recent detectors such as YOLOv8 have shown promising results, their backbone architectures lack explicit mechanisms to guide multi-scale feature refinement, limiting performance on high-resolution aerial data. In this work, we propose YOLO-SPCI, an attention-enhanced detection framework that introduces a lightweight Selective-Perspective-Class Integration (SPCI) module to improve feature representation. The SPCI module integrates three components: a Selective Stream Gate (SSG) for adaptive regulation of global feature flow, a Perspective Fusion Module (PFM) for context-aware multi-scale integration, and a Class Discrimination Module (CDM) to enhance inter-class separability. We embed two SPCI blocks into the P3 and P5 stages of the YOLOv8 backbone, enabling effective refinement while preserving compatibility with the original neck and head. Experiments on the NWPU VHR-10 dataset demonstrate that YOLO-SPCI achieves superior performance compared to state-of-the-art detectors. |
| title | YOLO-SPCI: Enhancing Remote Sensing Object Detection via Selective-Perspective-Class Integration |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2505.21370 |