[IET Image Processing 2022] MobileTrack: Siamese efficient mobile network for high-speed UAV tracking

This is an official pytorch implementation of the 2022 IET Image Processing paper:

MobileTrack: Siamese efficient mobile network for high-speed UAV tracking
(accepted by IET Image Processing, DOI: 10.1049/ipr2.12565)

The paper can be downloaded from IET Image Processing

The models and raw results can be downloaded from BaiduYun.

UAV Tracking

Datasets	mobiletrack_r50_l234
UAV123(Suc./Pre.)	0.609/0.813
UAVDT(Suc./Pre.)	0.559/0.774
DTB70(Suc./Pre.)	0.612/0.814

Note:

• r50_lxyz denotes the outputs of stage x, y, and z in ResNet-50.
• The suffixes DTB70 is designed for the DTB70, the default (without suffix) is designed for UAV20L and UAVDT.
• e20 in parentheses means checkpoint_e20.pth

Installation

Please find installation instructions in INSTALL.md.

Quick Start: Using MobileTrack

Add SmallTrack to your PYTHONPATH

export PYTHONPATH=/path/to/mobiletrack:$PYTHONPATH

demo

python tools/demo.py \
    --config experiments/siamban_mobilev2_l234/config.yaml \
    --snapshot experiments/siamban_mobilev2_l234/MobileTrack.pth
    --video demo/bag.avi

Download testing datasets

Download datasets and put them into testing_dataset directory. Jsons of commonly used datasets can be downloaded from Google Drive or BaiduYun. If you want to test tracker on new dataset, please refer to pysot-toolkit to setting testing_dataset.

Test tracker

cd experiments/siamban_mobilev2_l234
python -u ../../tools/test.py 	\
	--snapshot MobileTrack.pth 	\ # model path
	--dataset UAV123 	\ # dataset name
	--config config.yaml	  # config file

The testing results will in the current directory(results/dataset/model_name/)

Eval tracker

assume still in experiments/siamban_mobilev2_l234

python ../../tools/eval.py 	 \
	--tracker_path ./results \ # result path
	--dataset UAV123         \ # dataset name
	--num 1 		 \ # number thread to eval
	--tracker_prefix 'ch*'   # tracker_name

Training 🔧

See TRAIN.md for detailed instruction.

Acknowledgement

The code based on the PySOT , SiamBAN , MobileNetV2 and ECA-Net We would like to express our sincere thanks to the contributors.

Citation:

If you find this work useful for your research, please cite the following papers:

@article{https://doi.org/10.1049/ipr2.12565,
author = {Xue, Yuanliang and Jin, Guodong and Shen, Tao and Tan, Lining and Yang, Jing and Hou, Xiaohan},
title = {MobileTrack: Siamese efficient mobile network for high-speed UAV tracking},
journal = {IET Image Processing},
volume = {16},
number = {12},
pages = {3300-3313},
doi = {https://doi.org/10.1049/ipr2.12565},
url = {https://ietresearch.onlinelibrary.wiley.com/doi/abs/10.1049/ipr2.12565},
eprint = {https://ietresearch.onlinelibrary.wiley.com/doi/pdf/10.1049/ipr2.12565},
abstract = {Abstract Recently, Siamese-based trackers have drawn amounts of attention in visual tracking field because of their excellent performance. However, visual object tracking on Unmanned Aerial Vehicles platform encounters difficulties under circumstances such as small objects and similar objects interference. Most existing tracking methods for aerial tracking adopt deeper networks or inefficient policies to promote performance, but most trackers can hardly meet real-time requirements on mobile platforms with limited computing resources. Thus, in this work, an efficient and lightweight siamese tracker (MobileTrack) is proposed for high-time Unmanned Aerial Vehicles tracking, realising the balance between performance and speed. Firstly, a lightweight convolutional network (D-MobileNet) is designed to enhance the characterisation ability of small objects. Secondly, an efficient object-aware module is proposed for local cross-channel information exchange, enhancing the feature information of the tracking object. Besides, an anchor-free region proposal network is introduced to predict the object pixel by pixel. Finally, deep and shallow feature information is fully utilised by cascading multiple anchor-free region proposal networks for accurate locating and robust tracking. Extensive experiments on the three Unmanned Aerial Vehicles benchmarks show that the proposed tracker achieves outstanding performance while keeping a beyond-real-time speed.},
year = {2022}
}

If you have any questions about this work, please contact with me via xyl_507@outlook.com