🛡️ RealityGuard-India

AI-Powered Misinformation Detection & Content Verification for the Indian Digital Ecosystem.

📌 Overview

Misinformation in India often spreads through visual media—news clippings, social media posts, and viral WhatsApp images. RealityGuard-India is a specialized tool designed to verify content by extracting text from images (OCR) and performing deep semantic analysis using Google Gemini 1.5 Flash to flag potential misinformation, propaganda, or deep-fake narratives.

🚀 Key Features

• Visual Intelligence: Uses advanced OCR to pull text from low-quality viral images.
• Contextual Analysis: Leverages Gemini's reasoning to understand cultural nuances and regional context specific to India.
• Risk Triage: Automatically classifies content into: Verified, Partially False, Misleading, or False.
• Streamlit Interface: A clean, user-friendly dashboard for real-time verification.

🛠️ Technical Stack

• Backend: Python (Streamlit)
• AI/ML: Google Gemini 1.5 Flash (NLP & Reasoning)
• Vision: Tesseract OCR (Image Processing)

📂 Project Structure

├── app.py                # Main Streamlit UI
├── ocr_engine.py         # Image-to-text processing pipeline
├── gemini_classifier.py  # AI reasoning and truth-checking logic
├── requirements.txt      # Dependency management
└── .env.example          # Template for API keys