xuqingyu26 / hgtmt

PyTorch implementation of "Heterogeneous Graph Transformer for Multiple Tiny Object Tracking in RGB-T Videos", IEEE Transactions on MultiMedia.

HGTMT

PyTorch implementation of "Heterogeneous Graph Transformer for Multiple Tiny Object Tracking in RGB-T Videos", IEEE Transactions on MultiMedia. Feel free to contact me ([email protected]) if any questions. Please star the repository~~

Dataset: VT-Tiny-MOT

• Download the dataset from Baidu Drive (Key: VTMT) and unzip them to ./data

Note:

• The multiple tiny object tracking dataset is composed of two modalities: visible and thermal, and is well aligned.

Code:

• Create the working environment through environment.yml
• Download the transformer pvtv2 backbone from PVTv2.
• Run training/main_RGBT-Tiny_graph_gnnloss.py for training.
• Run tracking/RGBT-Tiny_private_graph_Track2_crossmodal.py for tracking.

Acknowledgement:

• This repository benefits a lot from TransCenter and GSDT.

Cite:

@article{xu2024heterogeneous,
  title={Heterogeneous Graph Transformer for Multiple Tiny Object Tracking in RGB-T Videos},
  author={Xu, Qingyu and Wang, Longguang and Sheng, Weidong and Wang, Yingqian and Xiao, Chao and Ma, Chao and An, Wei},
  journal={IEEE Transactions on Multimedia},
  year={2024},
  publisher={IEEE}
}