📝 AI-Powered Handwritten Answer Sheet Evaluator

An AI-driven platform that automates the manual evaluation of handwritten answer sheets.This system enhances grading accuracy, fairness, and efficiency while offering powerful analytics and feedback capabilities.

🚀 Features

• ✍️ Handwritten Text Recognition using Google Vision OCR
• 🧠 Context-Aware Answer Evaluation powered by Gemini API (Semantic Analysis)
• 🔍 Synonym & Phrase Matching, Sentence Structure, and Logic Coherence Detection
• 📷 Mobile/Web Upload of Answer Sheets
• ➕ Mathematical Expression Parsing
• 📊 Performance Analytics & Adaptive Learning Insights
• 📥 Dynamic Feedback Reports for Students
• ☁️ Cloud-based Infrastructure with PostgreSQL & Real-time Processing

🖼️ System Architecture

User Upload (Web/Mobile) 
      │
      ▼
Preprocessing & Image Cleanup
      │
      ▼
Google Vision OCR → Extract Text
      │
      ▼
Gemini API → Semantic Evaluation
      │
      ▼
Intelligent Grading Logic (Keyword + Context)
      │
      ├─> Math Expression Recognition
      │
      ├─> Feedback Generator
      │
      ▼
Analytics Dashboard + Reports + PostgreSQL Storage

⚙️ Technologies Used

Tech	Purpose
Google Vision OCR	Handwritten Text Recognition
Gemini API (NLP)	Semantic Analysis of Answers
PostgreSQL	Secure Data Storage
NEXT js	Backend Server (API & Logic)
NEXT js	Frontend (Web / Mobile Interfaces)
NEXT js	RESTful API Layer

🛠️ Installation & Setup

1. Clone the Repository

git clone https://github.com/AKHIL-DyC/DeepGrade.git

2. Install Dependencies

npm install

3. Configure Environment Variables

GOOGLE_VISION_API_KEY=your_key_here
GEMINI_API_KEY=your_key_here
DATABASE_URL=postgres://...

4. Run the app

npm run dev

📄 Example Workflow

1. Teacher uploads the question paper and evaluation criteria via the portal.
2. Student uploads a scanned handwritten answer sheet through the web or mobile interface.
3. The system performs OCR using Google Vision to extract handwritten text.
4. Extracted answers are contextually evaluated using Gemini NLP, based on:
   • The uploaded question paper
   • The evaluation criteria
5. The system assigns marks considering semantic meaning, sentence structure, synonyms, and logic.
6. Personalized feedback and performance reports are generated for the student.
7. Educators and admins access detailed analytics to refine the evaluation.

✅ Future Improvements

• Voice-to-text feedback for accessibility.
• Multi-language answer sheet support.
• Educator AI assistant for instant doubt resolution.
• AI-based cheating/plagiarism detection.

🤝 Contributing

Pull requests are welcome! For major changes, please open an issue first to discuss what you would like to change.