Application to recognize and identify people based in a Convolutional Neural Network designed using TensorFlow.
To execute the application we have to follow the next steps
python src/take_face_square.py /<path>/<to>/<project>/data/shape_predictor_68_face_landmarks.dat
python src/create_convolutional_dataset.py /<path>/<to>/<project>/img
jupyter notebook
Execute notebook called notebooks/Convolutional Neural Network.ipynb
TurtleFlow technologies requirements:
- OpenCV - Open Source computer vision.
- Dlib - Machine Learning library.
- TensorFlow - TensorFlow is an Open Source Software Library for Machine Intelligence.
- Pandas - Data manage python library.
- Scikit-learn - Machine Learning library.
- Numpy - Data manage library.
Here you can find a pdf document with the full research process and project architecture. Full Documentation
Video tutorial explaining the project structure. Video Tutorial
Real demo identifying two different people. Demo