Skip to content

Latest commit

 

History

History
77 lines (46 loc) · 2.4 KB

README.md

File metadata and controls

77 lines (46 loc) · 2.4 KB

Test Portal using Face Recognition

This project is an online test portal that utilizes face recognition technology to authenticate users and provide a secure testing environment. The system identifies and verifies the test takers' faces to ensure the authenticity of the participants.

Features

  • Face recognition-based user authentication
  • Secure test environment
  • User registration and management
  • Test creation and management
  • Test-taking functionality
  • Scoring and result generation

Technologies Used

  • Python
  • Django (Python web framework)
  • OpenCV (Open Source Computer Vision Library)
  • CNN (Machine learning algorithm)
  • HTML, CSS, JavaScript (Front-end development)

Screenshots

Login Page

Face Recognition Page

Login Failed Alert

Home Page

Quiz Instructions

Quiz Questions

Installation

  1. Clone the repository:

    git clone https://github.com/akanksha2911/FiyePr.git
    
  2. Install the required dependencies:

    pip install -r requirements.txt
    
  3. Change the path of files as relative path is not specified.

  4. Set up the database:

    Create your own dataset using generate.py for each face. The generate.py will create folder of all faces which will be used for training the model. Update the ResultsMap.pkl and classifier_model.h5 file based on your trained dataset.

  5. Run the application:

    cd Quiz2
    python manage.py runserver
    
  6. Access the application in your browser at http://127.0.0.1:8000/.

Usage Steps

  1. Log in using your credentials.
  2. The system will use face recognition to verify your identity.
  3. Create a new test or select an existing one.
  4. Start the test and follow the instructions.
  5. Submit the test, and your score will be displayed.
  6. Being a superuser you can even add quiz, questions.

Contributing

Contributions are welcome! If you find any issues or have suggestions for improvements, please open an issue or submit a pull request.