Welcome to pyppbox !

This repo was reinitiated from version 1.0b9. The complete history is available here 33da563.

---

🐍📦 pyppbox = Python + People + Toolbox

• Yes, pyppbox is an open-source Python toolbox which was originally made for the PoseTReID framework.
• This toolbox features 3 main modules: People Detection module, People Tracking module, and People Re-identifying module. These modules support a bunch of well-known people detectors, trackers, and re-identificators which can be used separately or put together with only a few lines of code.
• pyppbox also supports real-time online and offline evaluation on PoseTReID datasets.
• The initial version of pyppbox is also intergrated with GUI for easy demo and config.

⚠️ IMPORTANT NOTICES (Febuary 10, 2023)

• 🆕 Introduce GLOBAL & LOCAL cfg for multi-threading application -> Check examples page for more details.
• Add GPLv3+ license
• OpenPose submodule is completely removed due to the complicated license of OpenPose and it will be replaced by MMPose in the future versions. Stay tune for this amazing update! 🆕 Maybe the official YOLOv7 instead! 🆕
• Please refer to this repo pyppbox-paper for those who need OpenPose or need to reproduce the results in the paper.

⚽ Comparison Results

• Comparisions on PoseTReID datasets (Check our pre-printed paper: http://arxiv.org/abs/2205.10086)
• Click here for raw results in the paper

⚙️ Requirements

• Please check the README.md in requirements (Can be installed later).

🚀 Setup pyppbox

Option 1: Use the prebuilt WHL file here

• pip install pyppbox-xxx.whl

Option 2: Build your own pyppbox

• Download the extra models & weights and extract to the root pyppbox
• Create WHL by running creat_whl.cmd
• Install newly created WHL pip install pyppbox-xxx.whl

Quick Test

• On your terminal or CMD:

python
import pyppbox
pyppbox.launchGUI()

• Now you should see the GUI of pyppbox for easy demo. Just hit the "LAUNCH" button!

📝 Documentation?

• COMING SOON ⌛
• Meanwhile you can check the examples here !

🔗 Citation

• Extension of PoseTReID paper (Pre-printed on ARXIV):

@INPROCEEDINGS{ptreid9946587,
  author={Siv, Ratha and Mancas, Matei and Gosselin, Bernard and Valy, Dona and Sreng, Sokchenda},
  booktitle={2022 9th International Conference on Electrical Engineering, Computer Science and Informatics (EECSI)}, 
  title={People Tracking and Re-Identifying in Distributed Contexts: Extension Study of PoseTReID}, 
  year={2022},
  volume={},
  number={},
  pages={337-342},
  doi={10.23919/EECSI56542.2022.9946587}}

• Original PoseTReID paper:

@INPROCEEDINGS{ptreid9271712,
  author={Siv, Ratha and Mancas, Matei and Sreng, Sokchenda and Chhun, Sophea and Gosselin, Bernard},
  booktitle={2020 12th International Conference on Information Technology and Electrical Engineering (ICITEE)}, 
  title={People Tracking and Re-Identifying in Distributed Contexts: PoseTReID Framework and Dataset}, 
  year={2020},
  pages={323-328},
  doi={10.1109/ICITEE49829.2020.9271712}}