MoveNet_Yoga

This project identifies and evaluates several yoga movements.

Related Library

• Python (programming language):

• The PyPA recommended tool for installing Python packages:

• Fundamental package for array computing in Python:

• Powerful data structures for data analysis, time series, and statistics:

• Fast, Extensible Progress Meter:

• Pure python download utility:

• Source machine learning framework for everyone:

• Python plotting package:

• Wrapper package for OpenCV python bindings:

• A set of python modules for machine learning and data mining:

• Parser for command-line options, arguments and sub-commands:

Recommended IDE Setup

VSCode + Python

Project Setup

Create a virtual environment for the project

Virtual environment venv

python -m venv venv

Activate virtual environment:

venv\Scripts\activate

Update to the latest pip version:

python.exe -m pip install --upgrade pip

- Install the external dependencies needed for the project:

pip install -r setup.txt

Make data

Skeleton is extracted like below FIG

python make_csv.py --source yoga_cg --data data

Training

Make model .h5 to detect yoga poses

python train.py --train data/train_data.csv --test data/test_data.csv --output models/model_yoga_LSTM.h5 --epochs 200 --batch 64 --patience 20

Inference detection

Detect several pose of yoga include: chair, cobra, dog, tree, and warrior

python test.py --model models/model_yoga_LSTM.h5 --data images/tree3.jpg

Deactivate env

Once you’re done working with this virtual environment, you can deactivate it:

deactivate