This is a Generic Object Tracking Project.
- Improving Visual Object Tracking through Visual Prompting | Code available!
This is the official repository for "Improving Visual Object Tracking through Visual Prompting."
The raw results can be downloaded from here.
Dataset | Model | AUC | OP50 | OP75 | Precision | NPr |
---|---|---|---|---|---|---|
NfS-30 | ToMP-50 | 66.86 | 84.36 | 53.50 | 80.58 | 84.00 |
PiVOT-L-27 | 68.22 | 86.05 | 55.45 | 84.53 | 86.66 | |
OTB-100 | ToMP-50 | 70.07 | 87.83 | 57.79 | 90.83 | 85.98 |
PiVOT-L-27 | 71.20 | 89.35 | 55.73 | 94.58 | 88.46 | |
UAV123 | ToMP-50 | 68.97 | 83.84 | 64.63 | 89.70 | 84.79 |
PiVOT-L-27 | 70.66 | 85.69 | 67.06 | 91.80 | 86.74 | |
LaSOT | ToMP-50 | 67.57 | 79.79 | 65.06 | 72.24 | 77.98 |
PiVOT-L-27 | 73.37 | 85.64 | 75.18 | 82.09 | 84.68 | |
AVisT | ToMP-50 | 51.61 | 59.47 | 38.88 | 47.74 | 66.66 |
PiVOT-L-27 | 62.18 | 73.25 | 55.46 | 65.55 | 81.20 |
The codebase is built based on PyTracking.
Familiarity with the PyTracking codebase will help in understanding the structure of this project.
git clone https://github.com/chenshihfang/got-pivot.git
Ensure that CUDA 11.7 is installed.
sudo apt-get install libturbojpeg
Run the installation script to install all the dependencies. You need to provide the conda install path and the name for the created conda environment
bash install_PiVOT.sh /your_anaconda3_path/ got_pivot
conda activate got_pivot
You can follow the setup instructions from PyTracking.
There are two different local.py
files located in:
ltr/admin
pytracking/evaluation
python evaluate_PiVOT_results.py
The pretrained model can be downloaded from here.
-
First, set the parameter
self.infer
toTrue
inltr/models/tracking/tompnet.py
. -
Second, set up the Pretrained Model path in
pytracking/pytracking/parameter/tomp/pivotL27.py
. -
Then execute the following command:
CUDA_VISIBLE_DEVICES=0 python pytracking/run_experiment.py myexperiments_pivot pivot --debug 0 --threads 1
-
First, set the parameter
self.infer
toFalse
in:ltr/models/tracking/tompnet.py
-
Then, proceed with the following stages:
Stage 1:
python ltr/run_training.py tomp tomp_L_27
Stage 2: Place the
tomp_L_27
checkpoint in:ltr/train_settings/tomp/pivot_L_27.py
Then run:
python ltr/run_training.py tomp pivot_L_27
This codebase is implemented on PyTracking libraries.
If you find this repository useful, please consider giving a star ⭐ and citation 👍
@article{pivot2024tmm,
title={Improving Visual Object Tracking through Visual Prompting},
author={Chen, Shih-Fang and Chen, Jun-Cheng and Jhuo, I-Hong and Lin, Yen-Yu},
journal={IEEE Trans. Multimedia (TMM)},
year={2024},
publisher={IEEE}
}