This code has been tested on Ubuntu 18.04, Python 3.8.3, Pytorch 0.7.0/1.6.0, CUDA 10.2. Please install related libraries before running this code:

pip install -r requirements.txt

Download pretrained model and put it into tools/snapshot directory.

Download testing datasets and put them into test_dataset directory. If you want to test the tracker on a new dataset, please refer to pysot-toolkit to set test_dataset.

python test.py 	                          \
	--trackername SiamAPN           \ # tracker_name
	--dataset UAV10fps                  \ # dataset_name
	--snapshot snapshot/general_model.pth   # model_path

The testing result will be saved in the results/dataset_name/tracker_name directory.

Download the datasets：

• VID
• YOUTUBEBB (code: t7j8)
• COCO
• GOT-10K

Note: train_dataset/dataset_name/readme.md has listed detailed operations about how to generate training datasets.

To train the SiamAPN model, run train.py with the desired configs:

cd tools
python train_apn.py 

The pre-trained model can be found at (epoch=37) : general_model(code:w3u5)

We provide the tracking results_v1 (code: s3p1) of UAV123@10fps, UAV20L, and VisDrone2018-SOT-test. Besides, the tracking results_v2 (code: j4t5) of UAV123@10fps, UAV20L, VisDrone2018-SOT-test and UAVTrack112 are also provided.

The pre-trained model can be found at (epoch=25): general_model(code:j29k)

We provide the tracking results (code: xb41) of UAV123@10fps, UAV20L.

**Note:**The pre-trained model of SiamAPN and SiamAPN++ can also be found at googledriver

If you want to evaluate the tracker mentioned above, please put those results into results directory.

python eval.py 	                          \
	--tracker_path ./results          \ # result path
	--dataset UAV10fps                  \ # dataset_name
	--tracker_prefix 'general_model'   # tracker_name

UAVTrack112 benchmark is created from images captured during the real-world tests. It can be downloaded at UAVTrack112 (code: xb41). If you want to use this UAVTracking benchmark, please cite the paper below. More detail of the benchmark are available in here.

@INPROCEEDINGS{fu2021siamese,       
	author={Fu, Changhong and Cao, Ziang and Li, Yiming and Ye, Junjie and Feng, Chen},   
	booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)}, 
	title={{Siamese Anchor Proposal Network for High-Speed Aerial Tracking}},
	year={2021},
	volume={},
	number={},
	pages={1-7}
}

@INPROCEEDINGS{fu2021tgrs,       
	author={Fu, Changhong and Cao, Ziang and Li, Yiming and Ye, Junjie and Feng, Chen},   
	journal={IEEE Transactions on Geoscience and Remote Sensing}, 
	title={{Onboard Real-Time Aerial Tracking with Efficient Siamese Anchor Proposal Network}},
	year={2021},
	volume={},
	number={},
	pages={1-13}
}

If you have any questions, please contact me.

Ziang Cao

Email: [email protected]

The code is implemented based on pysot. We would like to express our sincere thanks to the contributors.