1. Introduction
2. Problem Statement
3. Solution Overview
4. Tech Stack
5. Architecture / Diagram (if any)
6. Installation & Usage

(Dyslexia), a common learning difficulty affecting 15–20% of the global population, often goes undiagnosed until it significantly impacts academic and emotional development. Many students face delayed diagnosis due to the lack of accessible tools, leading to missed opportunities for timely intervention.

Our project presents an advanced educational tool that leverages eye-tracking and audio analysis to detect and support students with dyslexia and other learning disabilities. It builds a complete ecosystem connecting students, teachers, and support professionals to ensure early identification and intervention.

• Prevalence: (1 in 5 students) are affected by dyslexia, often resulting in lower academic performance and reduced self-esteem.
• Detection Gap: Traditional methods miss up to 50% of dyslexic students (NIH), delaying necessary support.
• Delayed Intervention: Without early detection, dropout rates increase drastically. With intervention, dropout rates can be reduced by 70% and lifetime earnings boosted by 20% (Yale Center for Dyslexia & Creativity).
• Lack of Tools: Current systems lack personalization, early screening access, and real-time progress tracking.

We propose a unified platform that:

1. 🧾 Parent-Friendly Assessment
   A smart, form-based system where parents can answer structured questions to auto-generate an initial risk report.

2. 🔍 AI-Powered Detection Engine
   Uses a combination of eye-tracking, audio pattern analysis (Librosa), and facial expression detection to assess dyslexia likelihood, delivering detailed visual reports.

3. 🧠 Cognitive Skill-Building Quizzes
   Dynamically generated quizzes aligned with a student’s performance metrics to boost memory, focus, and comprehension.

4. 🎮 Gamified Learning Modules
   Engaging and accessible games designed to strengthen phonological awareness and reading fluency.

5. 🤖 AI Chatbot for Learning Assistance
   An intelligent chatbot to guide students through lessons, answer questions, and recommend activities based on weaknesses detected.

What makes us unique:

• Multimodal Analysis: Unlike traditional screeners, we analyze visual, auditory, and emotional cues in real time.
• End-to-End Support: From early screening to personalized remediation, everything is integrated into one ecosystem.
• Inclusive & Accessible: Designed for both parents and educators — no technical expertise required.
• Research-Backed Accuracy: Eye-tracking models reach up to 85% detection accuracy, surpassing traditional methods (Source).
• Continuous Learning Loop: The platform evolves with the user through performance feedback and AI recommendations.

• Frontend: ReactJS, TypeScript, TailwindCSS
• Backend: FastAPI (Python)
• Database: MongoDB
• Core Logic & ML: OpenCV, Mediapipe
• Tools: GitHub, VSCode
• Hosting: Vercel, Render

📌 This includes Eye Tracking Module, Audio Analysis Engine, ML-Based Analyzer, Report Generator, and Web/App Interfaces.

• Docker
• Python (FastAPI)
• TypeScript
• MongoDB
• OpenCV
• Mediapipe
• Libraries from requirements.txt and package.json

## Clone the repository
git clone https://github.com/FantomCode25/Innov8ors.git

# Backend Setup
cd Innov8ors
cd backend

## Build and run with Docker

docker build -t backend-api .
docker run -p 8000:8000 backend-api

pip install -r requirements.txt
python app.py

## To run without Docker:

## Create virtual environment
python -m venv venv
source venv\Scripts\activate

## Install dependencies
pip install -r requirements.txt

## Run the application
uvicorn main:app --reload


# Frontend Setup (Web)
cd Innov8ors
cd frontend

# Install dependencies
npm install

# Start the development server
npm run dev

We have well maintained proper backend documentation After cloning Access the API at http://localhost:8000/docs Our Hosted API DOCS: https://dylexiaai.onrender.com/docs and https://dylexiaai.onrender.com/redoc

Our Hosted Project: Frontend: https://dylexia.vercel.app/ Backend: https://dylexiaai.onrender.com/

💡 Each member contributed to ideation, testing, and final presentation.