This repository contains the official implementation of our paper: MAC-DETR: Mamba for Robust Object Detection in High-Resolution Remote Sensing

📄 The full paper will be available upon publication. This repository will be updated with the link.

Our model is evaluated on three public datasets. Please follow the instructions below to prepare them.

1.HRSC2016

Description: A dataset for ship detection in aerial imagery.

Download: Obtain the dataset from IEEE DataPort - HRSC2016.

2.DOTA

Description: A large-scale dataset for object detection in aerial images with oriented bounding boxes (OBB).

Download: Clone the repository from Aerial-YOLO-DOTA or visit the official DOTA website for the original data.

3.NWPU-VHR-10

Description: A ten-class object detection dataset for very high-resolution remote sensing images.

Download: Clone the repository from NWPU-VHR-10.

Performance results on various remote sensing datasets. TBD (To Be Determined) values will be updated upon completion of training.