/yolo-hand-pose

Primary LanguageJupyter NotebookApache License 2.0Apache-2.0

YOLOv8n-Pose Hand Pose Detection

results

This repository contains the code and resources for a custom hand pose detection model trained using the YOLOv8n-pose framework by ultralytics. The model is trained on a custom dataset of hand keypoints available on Kaggle.

Important Note: Due to computational limitations, the model has only been trained to 50% of its potential. However, it still demonstrates excellent performance.

Project Goal

The primary objective of this project is to contribute to the development of state-of-the-art (SOTA) models for hand pose detection, specifically targeting applications in sign language classification and AR/VR.

What's Included:

  • Training scripts for the YOLOv8n-pose model on the custom hand keypoint dataset.
cd ./train
  • Pre-trained model weights (50% trained).
cd ./model
  • Configuration files and scripts for inference with the trained model.
cd ./inference

Future Developments:

  • Continue training the model to achieve full capacity and improve accuracy.
  • Explore advanced techniques for hand pose estimation and landmark detection.
  • Integrate the model into sign language classification and AR/VR projects.

Contribution

We welcome contributions from the community to enhance this project. Feel free to:

  • Fork the repository and experiment with different training configurations.
  • Implement additional functionalities for hand pose estimation.
  • Share your ideas for integrating the model into sign language and AR/VR applications.

Let's build better hand pose detection models together!

training_graph


Follow me

GitHub @RionDsilvaCS · Linkedin @Rion Dsilva · Twitter @Rion_Dsilva_CS