Multi-Drone-Multi-Object-Detection-and-Tracking

Installation

1.create a conda virtual environment and activate it:

conda create -n mdmt python=3.8
conda activate mdmt

2.install pytorch and suited mmcv-full, please refer to .

(if you have no idea which version to install, stay with ours: torch 1.10.0+cu113/ torchvision 0.11.1+cu113/ mmcv-full 1.5.0/ mmdet 2.25.1)

3.download the codes and :

cd Multi-Drone-Multi-Object-Detection-and-Tracking-main
pip install -r requirements.txt
cd mmdet
pip install -r requirements/build.txt
pip install -v -e .  # or "python setup.py develop"
cd ..
pip install -r requirements.txt
pip install -v -e .  # or "python setup.py develop"

Getting started

Dataset Structure

Multi-Drone-Multi-Object-Detection-and-Tracking-main

data

MDMT

train/ val/ test/

Inference

1.Inference MIA-Net:

python ./demo/supplement_MIA.py

import arguments:

--config config file

--input input data folder

--xml_dir input xml file of the groundtruth

--result_dir the directory to save results, no "/" in the end

--method the sub-directory used in result_dir, representing different methods

2.Inference MIA-Net(w/o supplementation), MIA-Net(w/ localmatching), MIA-Net(w/ globalmatching), run:

python ./demo/multiDrone_matchingIDallocation-NMS.py
python ./demo/multiDrone_localmatching-NMS.py
python ./demo/multiDrone_globalmatching-NMS.py

Evaluation

1.For tracking perforcement:

python ./demo/eval/json_2_txt.py [--sequences_result] [--output_dir]
python ./demo/eval/txttxt_test.py [--test_file_dir]

2.For MDA score:

python ./demo/eval/mango_eval.py [--sequences_result]

Citation

@article{liu2023robust,
    title={Robust Multi-Drone Multi-Target Tracking to Resolve Target Occlusion: A Benchmark},
    author={Liu, Zhihao and Shang, Yuanyuan and Li, Timing and Chen, Guanlin and Wang, Yu and Hu, Qinghua and Zhu, Pengfei},
    journal={IEEE Transactions on Multimedia},
    year={2023},
    publisher={IEEE}
}