📚 Table of Contents

• About
• Get Started
• To-Do
• Contribute

📌 About

This is an low-code / no-code application that dramatically reduces the complexity of using pose estimation frameworks to extract pose keypoints/landmarks. It will support two libraries for pose keypoint extraction. ( MMPose, MediaPipePoseModel).

This application is intended for anyone (commercial, research, non-profit, hobbyist) who wants to easily ultilize common pose estimation frameworks. Supports Windows and Linux.

Example Usage (WIP)

CLI:

"win_keypoint_extraction_cli.exe -input_path input_path -output_path output_path -pose_framework mediapipe"

📈 Get Started

No prereqs required.

NOTE: Application yet to be implemented.

CLI:

1. Download the CLI application for your OS

2. Use the app in your terminal with the following arguments:

   Required:

   1. *-input_path* Path to your input directory or file. Pass "webcam" or webcam device number to use a webcam.
   2. *-output_path* Path to your output directory.
   3. *-pose_framework* Name of the framework you are using. Future supported options: "mediapipe" or "mmpose"
   

   Optional:

   4. *--pose_config* Optionally specify path to a model configuration file
   5. *--pose_checkpoint* Optionally specify poth to model checkpoint file.
   6. *--detection_framework* Optionally specify name of detection framework to use
   7. *--detection_config* Optionally specify path to detection model configuration file.
   8. *--detection_checkpoint* Optionally specify path to detection model checkpoint file.
   

3. Pour yourself a cup of coffee while you wait for your results.

GUI (WIP):

Still Needed

Test Data

1. Images: https://github.com/jeffffffli/CrowdPose
2. Videos: Still needed

🎯 To-Do

• Finish Todo List (#1)
• Make Working Demo (#2)

🙋 Contribute

We welcome contributions from everyone! Whether you're fixing bugs, adding features, improving documentation, or just asking questions - your help makes this project better.