A Multi-level Information Injection and Wavelet Sampling Learning Network(MWLN) for Aerial Imagery Object Detection is proposed.
git clone https://github.com/mi-luo/MWLN # clone
cd MWLN
pip install -r requirements.txt # installOur project is developed based on the ultralytics-yolov8 project, and its training, validation, and detection procedures are identical to those of the yolov8 project. For detailed operations, please refer to the following link: https://docs.ultralytics.com/#yolo-a-brief-history.
This paper and the related experiments utilize the publicly available datasets DOIR and NWPU VHR-10, with the code architecture developed based on YOLOv8. Below are the relevant paper titles associated with the experimental data and code. 1.“Object detection in optical remote sensing images: A survey and A new benchmark“ 2.“A Survey on Object Detection in Optical Remote Sensing Images“ 3.“Real-time flying object detection with yolov8“
We sincerely appreciate the spirit of open sharing demonstrated by these organizations and hope that their datasets and codes can provide valuable insights for research and development in the field of high-resolution remote sensing.
NOTE：Please note that the listing above is not in any particular order
For MWLN bug reports and feature requests, and feel free to ask questions and engage in discussions!