Correlation Filters for Unmanned Aerial Vehicle-Based Aerial Tracking: A Review and Experimental Evaluation

Changhong Fu, Bowen Li, Fangqiang Ding, Fuling Lin, and Geng Lu

This code library gives our experimental results and most of the publicly available DCF-based trackers.

The trackers are in folder tracker_set and the results are in all_trk_result.

In folder demo, we offer an example on how to utilize the trackers in this code library and how to implement the tracking procedure.

The submitted version of our paper has been uploaded to arxiv which can be found at:

https://arxiv.org/abs/2010.06255.

If you want to use our experimental results or related content, please cite our paper using the format as follows:

@ARTICLE{9445732,

author={Fu, Changhong and Li, Bowen and Ding, Fangqiang and Lin, Fuling and Lu, Geng},

journal={IEEE Geoscience and Remote Sensing Magazine},

title={Correlation Filters for Unmanned Aerial Vehicle-Based Aerial Tracking: A Review and Experimental Evaluation},

year={2021},

volume={},

number={},

pages={2-387},

doi={10.1109/MGRS.2021.3072992}

}

Results_OPE_win10_i7_2080_R2019a

The trackers are tested on the following platform.

  • windows 10 64-bit
  • Intel Core i7-8700K(3.70GHz)
  • 32G RAM
  • Nvidia GeForce RTX 2080
  • Matlab 2019a
  • CUDA10
  • VS2017

All the DCF-based trackers' results are obtained using a single CPU with a single core.

fig

Here shows some of the tracking results of 21 handcrafted DCF-based trackers.

DTB70

UAV123@10fps

UAVDT

VisDrone2019-SOT

UAVDT_Comparision against deep trackers

The deep trackers used GPU acceleration, while the DCF-based tracker AutoTrack used a single CPU only.

Acknowledgements

We thank the contribution of Siyi Li, Matthias Muller, Dawei Du, and Pengfei Zhu for their previous work of the benchmarks DTB70, UAV123, UAV123@10fps, UAV20L, UAVDT, and VisDrone2019-SOT. Their papers and benchmark address are listed here.

DTB70

paper: Visual object tracking for unmanned aerial vehicles: a benchmark and new motion models

paper site: https://dl.acm.org/doi/10.5555/3298023.3298169

code and benchmark site: https://github.com/flyers/drone-tracking

UAV123 & UAV123@10fps &UAV20L

paper: A Benchmark and Simulator for UAV Tracking

paper site: https://link.springer.com/chapter/10.1007%2F978-3-319-46448-0_27

code and benchmark site: https://cemse.kaust.edu.sa/ivul/uav123

UAVDT

paper: The Unmanned Aerial Vehicle Benchmark: Object Detection and Tracking

paper site: https://link.springer.com/article/10.1007/s11263-019-01266-1

code and benchmark site: https://sites.google.com/site/daviddo0323/projects/uavdt

VisDrone2019-SOT

paper: VisDrone-SOT2019: The Vision Meets Drone Single Object Tracking Challenge Results

paper site: https://ieeexplore.ieee.org/document/9022632

code and benchmark site: http://www.aiskyeye.com/