📚 pose-format Tutorials

Practical, end-to-end notebooks and examples that teach the pose-format library from basics to advanced workflows (conversion, Normalizing Data,Interpolation visualization, augmentation, and web usage). Written for researchers and developers working with sign language and human pose data.

• Related project: pose — alternative/related pose repository for SL processing.
• Documentation: pose-format Documentation — full pose-format API docs and guides.
• Paper: pose-format Paper — academic reference relevant to pose-format research.
• Editor extension (optional): For VS Code users, we recommend the Pose extension: Pose Extension — adds syntax highlighting and helpers for pose files.

All notebooks include an Open in Colab badge at the top and a one-line setup cell that installs the lesson-specific packages.

#	Notebook	Colab	What you learn
01	01_extract_landmarks_from_video.ipynb		Extract landmarks from video with MediaPipe Holistic and save .pose files
02	02_convert_pose_formats.ipynb		Convert .pose to JSON/NPZ and back
03	03_read_pose_files.ipynb		Load and slice pose data, work with frames and time ranges
04	04_visualize_pose.ipynb		Render videos, GIFs, and images from pose data
05	05_Normalization.ipynb		Normalize pose data for consistency across videos
06	06_Augmentation.ipynb		Apply data augmentation techniques (rotation, zoom, skew, etc.)
07	07_Interpolate.ipynb		Interpolate missing frames and smooth pose sequences
08	08_advanced_features.ipynb		Advanced features and techniques for pose data manipulation

• Python 3.11
• ffmpeg available on PATH for video handling (recommended)
• GPU optional for deep learning examples

1. Create and activate a conda environment. Install Python package manager (Miniconda) skip this step if you already have it installed.

• Download and install MiniConda from here

1. Open a terminal and run the following commands:

git clone https://github.com/24-mohamedyehia/pose-format-tutorials.git
cd pose-format-tutorials
conda create -n pose-format-tutorials python=3.11 -y
conda activate pose-format-tutorials
python -m pip install -r requirements.txt

1. Go to First Jupyter Notebook. Good luck and have fun! 🚀

Issues and pull requests are welcome to expand coverage, fix bugs, or improve examples.

This project is licensed under the MIT License - see the LICENSE file for details.