Official implementation of E.T.Track. E.T.Track utilized our proposed Exemplar Transformer, a transformer module utilizing a single instance level attention layer for realtime visual object tracking. E.T.Track is up to 8x faster than other transformer-based models, and consistently outperforms competing lightweight trackers that can operate in realtime on standard CPUs.

E.T.Track	The standard attention vs our Exemplar Attention on the right

Install the python environment using the environment file ettrack_env.yml.

Generate the relevant files:

python -c "from pytracking.evaluation.environment import create_default_local_file; create_default_local_file()"
python -c "from ltr.admin.environment import create_default_local_file; create_default_local_file()"

• Modify local.py. Modify the path files for the evaluation in pytracking/evaluation/local.py

We evaluate our models using PyTracking. The sequences comparing E.T.Track and LT-Mobile in the ''Video Visualizations'' section can be found here.

• Add the correct dataset in pytracking/experiments/myexperiments.py (default: OTB-100)
• Run python3 -m pytracking.run_experiment myexperiments et_tracker --threads 0

If you use this code, please consider citing the following paper:

@inproceedings{blatter2023efficient,
  title={Efficient visual tracking with exemplar transformers},
  author={Blatter, Philippe and Kanakis, Menelaos and Danelljan, Martin and Van Gool, Luc},
  booktitle={Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision},
  pages={1571--1581},
  year={2023}
}