A Facial Login Registration Implementation

FaceLogin is a login and registration app that uses facial recognition as a form of validation. The face recognition is done within the OpenCV framework in Python.

Setting up Project and Dependencies

1. Clone this repository on your local machine:

   git clone https://github.com/EmilChoparinov/FacialLogging

2. Enter the directory FacialLogging (cd ./FacialLogging)

Install npm packages

We must now install the servers and clients packages.

1. Navigate to "./server/intermediary/" and run the command: npm install
2. Navigate back to the root of the project by doing the command: ./../../../
3. Now install the client's packages by navigating to "./client/WebApp/" and running the command: npm install

Install python packages

We must now install the python dependencies in order to run the ml microservice. This was written in Python 3 so make sure Python 3 is install on your local machine

1. Navigate to: "./server/ml/"
2. Start a virtual environment in python3 by doing: py -3 -m venv FaceLoginEnv
3. Enter your virtual environment by running: . ./FaceLoginEnv/Scripts/activate
4. Install the requirements from r.txt using the command: pip install -r r.txt

Running the Project

This project requires three things to be run:

1. Navigate from the root directory back into "./client/WebApp/" and run the command ng build note you must have angular cli tools installed globally or locally. Currently the package name is @angular/cli.
2. Once that finishes building, navigate from the root directory into "./server/intermediary/" and run the command ts-node ./server.ts. If everything is successful a listening message should be displayed.
3. Navigate from the root directory into "./server/ml/" and run the command py -3 server.py.
4. If the project is all up and running successfully attempt going to localhost on port 8000 and try it out!

Architecture

• Implemented in the MEAN MVVC Architecture with a Flask microservice for OpenCV.
• Express is used as a proxy to the Python machine learning microservice that generates profiles and validates users. This architecture allows easy scalability due to how the microservice is setup.
• More info in this presentation here.

This chart represents the logic happening between Angular, the Express proxy, and Flask:

Screenshots

Contributors: @Ayadlin @Bhkangw @EmilChoparinov