Hackprague face-recognition
Face recognition app for marketing built during the HackPrague hackathon.
Face detection
Implementation of the face detector is in the folder face_detector.
It works based on the Viola Jones object detection framework
Example visualization
Face recognizer
Implementation of the face detector is available in the face_classifier folder.
Model is based on the pretarined VGG-16 neral network that is being fine-tuned for the desired faces.
Http Server
Using easy Flask server.
It can be run by python http_server/flask_server.py + you have to set the custom path to to your trained model in the server file.
Frontend
Implementation of the frontend is in the folder webcam and it's implemented in HTML and JavaScript.
You can run it by cd webcam and node index.js and opening the url from the terminal in your browser.
Example frontend screenshot
