Smart lock using facial recognition on top of RaspberryPi 3+ models

Script usage:

cam2sam.py: Collect samples from camera. cam2sam offers: - ID config - Face size and target size
camera.py: Labelize face using specific model.

What we are doing:

We researched for efficient algorithm for face classification, end up with SVM-RBF Steps to reproduce: Follow experiments.ipynb to execute tests and see SVM-RBF actual configs
We built client-server model for applying model into practice:

Setup:

pip3 install -r requirements.txt

cd server
python3 server.py

cd ../client
GPIO=true python3 client.py

Caution:

GPIO=true will require package RPi.GPIO to be installed
We offer config in client.py, please have a look at the script for proper config to make script works on your environment

Some screenshots:

Please leave us feedback if there are some issues with this repo. Enjoy your day!!!