AI EVM - (Electronic Voting Machine Powered by AI):

Demo

Click to watch full demo (by Satyendra Yadav)

Team

• Mohith Kune (me)
• Satyendra Yadav
• Venkatesh Vutukuri
• Ankit Chouhan
• Pooja Yadav

Introduction

Most of the sectors have been automated these days but the Voting process has stayed un-updated for years. We propose to leverage AI and use it with EVMs to bring revolution to the voting process.

The AI-powered EVMs can reduce many costs due to manual work at the polling booths. They also help to check malpractices by using Computer Vision.

Project Pre-requisite Steps

It is suggested to use virtual python==3.6 environment for this project.

If you have anaconda installed, you can create a new empty virtual enviroment as shown below

conda create --name AI_EVM_env python=3.6

# for activating env
conda activate AI_EVM_env

# for deactivating
conda deactivate

And navigate to the project to install dependecies.

(Please feel free to raise an issue if run into problems while installing dlib)

pip install -r requirements.txt

Now before running the Django project for the first time, db tables have to be created.

# makemigrations (It will generate schemas)
python manage.py makemigrations

# migrate (It will create tables in the db)
python manage.py migrate

# create an admin account for the project
python manage.py createsuperuser

Run the Django server on localhost on port 5000

python manage.py runserver 0.0.0.0:5000

Deployment

The project architecture is divided into three layers as below.

Since the server has to collect, analyze & reply with video stream in realtime, deploying server on internet at different location is not encourage.

The Presentation & Application layer are bind together in one Machine.

The main settings for deployment are:

• SECRET_KEY: It is important to make this variable secret. We propose to add this variable to the server environment & load at runtime as below.

# Read SECRET_KEY from an environment variable
import os
SECRET_KEY = os.environ['SECRET_KEY']

• DEBUG: This variable must be set to False before deploying!
• ALLOWED_HOSTS: This variable must contain the possible web domains from which we wish to visit the website.

ALLOWED_HOSTS = ['localhost', '0.0.0.0', '127.0.0.1', 'ai.evm.org']

• MailJet API Keys: These variable must be stored in the environment simialr to the SECRET_KEY

MAILJET_API_KEY = '*****************'
MAILJET_API_SECRET = '*****************'

• We propose to deploy using gunicorn & Nginx. The installation steps for these tools are different/specific to the platform we want to deploy on. Kindly refer the official documentation for installing.

• After installing gunicorn we can run the server using the below command

gunicorn --bind 0.0.0.0:8000 ai_evm.wsgi

• We can create a service/process on the server to run the above command & this step is also platform specific. For example we can create gunicron.service.

• If using Linux, we then submit this process to systemctl to run at every reboot.

sudo systemctl start gunicorn.service
sudo systemctl enable gunicorn.service
sudo systemctl status gunicorn.service

NOTE

• Face Recognition phase may not recognize your face
  • Head to models/src/Face_Recognition to generate files & place inside models/recognize_face_models.
• Create a superuser so that you can view admin section. The db included here has dummy data used for the demo.

Future Scope

We have added basic/major functionality. But always can be improved with more features like below.

• Biometric authentication can be added to strengthen auth process
• ATM like voting from remote places. No need to travel
• Online Voting is still a dream. We can try our best always!
• Voting data collected can be linked to aadhar, driving license, PAN etc - Govt db warehouse