Project Title : Eduproctor

Eduproctor is an innovative platform aimed at enhancing online learning experiences through real-time eye tracking and AI-generated quizzes.

Introduction

Eduproctor addresses the challenges faced by students in online learning environments, such as distractions and lack of engagement, by leveraging cutting-edge technologies. With real-time eye tracking and AI-generated quizzes, Eduproctor aims to improve student focus and comprehension during online classes.

Configuration & Installation

To get started with Eduproctor, ensure you have the following dependencies and tools installed:

Frontend

2. Next.js: Eduproctor is built using Next.js, a React framework for server-rendered applications along with webrtc . Install it globally using npm :

   npm install -g next
   

Once you have installed the necessary dependencies, follow these steps to set up the development environment:

1. Clone the Eduproctor repository from GitHub:

   git clone https://github.com/AkhileshJyotishi/EduProctor.git
   

2. Navigate to the project directory:

   cd EduProctor
   

3. Install project dependencies:

   npm install
   

4. Create a .env file in the root directory and add all environment variables.

   NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY=
   
   CLERK_SECRET_KEY=
   
   NEXT_PUBLIC_CLERK_SIGN_IN_URL=/sign-in
   
   NEXT_PUBLIC_CLERK_SIGN_UP_URL=/sign-up
   
   NEXT_PUBLIC_BASE_URL=http://localhost:3000
   
   NEXT_PUBLIC_STREAM_API_KEY=
   
   STREAM_SECRET_KEY=
   

5. Start the development server:

   npm run dev
   

6. Open your browser and navigate to http://localhost:3000 to access Eduproctor.

Backend

1. Download Python version 3.8 and it to your Environmental variables.

2. Set up virtual environment using following command :

   pip install virualenv
   
   cd backend
   
   cd QuestionGeneration
   
   python3.8 -m venv <virtual-environment-name>
   
   

3.    pip install flask
   
      pip install Flask-Cors
   
      pip install mediapipe
   

4. Run Following Commands

    python -m nltk.downloader universal_tagset
      
    python -m spacy download en
   
    pip install -U pip setuptools wheel
   
    pip install -U spacy
   
    python -m spacy download en_core_web_sm
   

5. You can download the weights from here:

wget https://github.com/explosion/sense2vec/releases/download/v1.0.0/s2v_reddit_2015_md.tar.gz

tar -xvf  s2v_reddit_2015_md.tar.gz

Details

Eduproctor offers a range of features designed to enhance the online learning experience:

1. Real-Time Eye Tracking:

• Eduproctor employs real-time eye tracking technology to monitor students' attention during online classes.

• If distractions are detected, alerts are sent to students, prompting them to refocus on the lesson.

2. AI-Generated Quizzes:

• The platform utilizes AI algorithms to generate personalized quizzes based on the lesson content.

• Quizzes are dynamically generated using keywords extracted from the material, ensuring relevance and alignment with lesson objectives.

3.Seamless Integration:

• Eduproctor seamlessly integrates with existing online learning platforms, providing a seamless experience for both students and educators.

• Teachers can easily incorporate Eduproctor into their online classes to enhance student engagement and comprehension.

Contributors

The following individuals have contributed to the EZCompose project:

1. Akhilesh Jyotishi

• Role: Frontend Developer
• Responsibilities: Did AI integration with frontend i.e. integrating realtime webrtc with tensorflow.js and mediapipe.

2. Yash Agarwal

• Role: Frontend Developer
• Responsibilities:Integrating AI quiz generator and realtime speech to text with Online Meet.

3. Darsh Baxi

• Role: AI-ML Engineer
• Responsibilities: MCQ Generation using T5-transformers.Extract keywords using PKE and mapping sentence with keywords using FlashText and generating wrong options for question using Sense2vec.

2. Samarth sahu

• Role: AI-ML Engineer
• Responsibilities: Assisted in developing logic and functions for detecting yawning and distractions using TensorFlow.js and MediaPipe.