Sovereign 8s.is mirror of standardgalactic/LMW-YOLO
- Python 99.6%
- Shell 0.2%
- HTML 0.1%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
| docker | ||
| docs | ||
| examples | ||
| tests | ||
| ultralytics | ||
| .dockerignore | ||
| .gitignore | ||
| CITATION.cff | ||
| CONTRIBUTING.md | ||
| LICENSE | ||
| mkdocs.yml | ||
| overall_architecture.png | ||
| pyproject.toml | ||
| README.md | ||
| requirements.txt | ||
| YOLO-UAV.yaml | ||
LMW-YOLO: Lightweight Model for Small Object Detection in Remote Sensing Images
This repository contains the official PyTorch implementation of the paper: "Lightweight model LMW-YOLO for small object detection in remote sensing images" (Under Review at Scientific Reports).
Please remember to add the datasets on your own before running the code.
🚀 Introduction
Detecting small objects in remote sensing images is challenging due to complex backgrounds and extreme scale variations. We propose LMW-YOLO, a lightweight object detection model. Key components include:
- LKCA (Large-Kernel Context Aggregation) module for shallow layers.
- MSDP (Multi-Scale Dilated Perception) module for deep layers.
- CSD (Context-Scale Decoupled) strategy.
- WIoU v3 loss function.
