This is the source code for this paper called "SiamBAG: Band Attention Grouping-based Siamese Object Tracking Network for Hyperspectral Videos"
Video tracking results in worker scenario:
worker.mp4
Note that: in this video, the red bounding box is SiamBAG tracker, the blue one is ground truth.
GIF tracking results in some scenarios:
 |
 |
 |
 |
 |
 |
Note that: in these scenarios, the black bounding box is SiamBAG tracker, the blue one is ground truth, and the red one is BAENet tracker.
Some important environments for SiamBAG.
- Python 3.9
- PyTorch 1.13.0
- CUDA 11.6
Please refer to 'SiamBAG-Installation_Environment.txt' file for more detailed.
Paper Download:
Pretrained model Download:
You can download the pretrained model.pth from Google Drive or Baidu Yun with extract password [gloh], and put the file under pretrained/siamrpn.
If these codes are helpful for you, please cite this paper:
BibTex Format:
@ARTICLE{10149343,
author={Li, Wei and Hou, Zengfu and Zhou, Jun and Tao, Ran},
journal={IEEE Transactions on Geoscience and Remote Sensing},
title={SiamBAG: Band Attention Grouping-based Siamese Object Tracking Network for Hyperspectral Videos},
year={2023},
volume={61},
number={},
pages={1-12},
doi={10.1109/TGRS.2023.3285802}}
Plain Text Format:
W. Li, Z. Hou, J. Zhou and R. Tao, "SiamBAG: Band Attention Grouping-Based Siamese Object Tracking Network for Hyperspectral Videos," in IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1-12, 2023, Art no. 5514712, doi: 10.1109/TGRS.2023.3285802.