README.md

One Homography is All You Need: IMM-JHSE

[Preprint Information Fusion 2024] One Homography is All You Need: IMM-based Joint Homography and Multiple Object State Estimation.

This repo was adapted from UCMCTrack. It is still under construction!

DanceTrack example:

Setting Up the Conda Environment

To create a new conda environment with the required dependencies, follow these steps:

1. Ensure you have Anaconda or Miniconda installed on your system.

2. Open a terminal or command prompt.

3. Navigate to the directory containing the environment.yml file.

4. Run the following command to create a new conda environment:

   conda env create -f environment.yml

5. Activate the newly created environment:

   conda activate <environment_name>

   Replace <environment_name> with the name specified in the environment.yml file.

Run IMM-JHSE on the DanceTrack test set

1. Download the DanceTrack testing and validation sets.

2. Extract the contents to maintain the following folder structure:

   {IMM-JHSE ROOT}
   |-- data
         |-- DanceTrack
         |   |-- val
         |   |   |-- dancetrack0004
         |   |   |   |-- img1
         |   |   |   |   |-- 00000001.jpg
         |   |   |   |   |-- ...
         |   |   |   |-- gt
         |   |   |   |   |-- gt.txt            
         |   |   |   |-- seqinfo.ini
         |   |   |-- ...
         |   |-- test
         |   |   |-- ...
   

3. Run the following to get results with video in the test_output/dance/test folder:

   python run_dance_test.py --seq all --param_file dancetrack_params.json --video

   You can omit the --video flag to perform inference without video output.

Test results

The result files used to obtain the results reported in the paper for the test sets are given in the result_files folder. Submit these to the evaluation servers for the various test datasets.

Citation

Please cite our article if you use this repo for further research:

@misc{claasen2024homographyneedimmbasedjoint,
      title={One Homography is All You Need: IMM-based Joint Homography and Multiple Object State Estimation}, 
      author={Paul Johannes Claasen and Johan Pieter de Villiers},
      year={2024},
      eprint={2409.02562},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2409.02562}, 
}