🚀 AI Scene Safety Analyzer

An advanced AI-powered application that analyzes images and determines safety levels using computer vision and large language models.

📌 Overview

The AI Scene Safety Analyzer automatically detects safety risks in images and classifies them into:

• ✅ Safe
• ⚠️ Warning
• 🔴 Dangerous

It also provides:

• Risk Score (0–100)
• Detailed explanation
• Recommended safety actions

---

🧠 How It Works

1. Image is uploaded via UI
2. BLIP Model generates image caption
3. Caption is sent to LLaMA 3.3-70B (via Groq API)
4. AI analyzes and returns:
   • Safety Label
   • Risk Score
   • Reason

---

✨ Features

• 🔍 AI-powered image safety analysis
• 🎯 Risk scoring system (0–100)
• 🌍 Multi-language support (7 languages)
• 🔊 Voice output (Text-to-Speech)
• 📊 Safety charts (Bar & Pie)
• 🖼️ Heatmap overlay on images
• 📄 Export reports:
  • PDF
  • CSV
  • Excel
• 💻 Beautiful UI using Gradio

---

🛠️ Technologies Used

• Python 3.12
• BLIP (Image Captioning Model)
• LLaMA 3.3-70B (via Groq API)
• Gradio (UI)
• PyTorch
• Pandas
• Matplotlib
• gTTS (Voice)
• deep-translator

---

🖥️ Project Structure

project/ │── app.py │── project.ipynb │── requirements.txt │── README.md