TrafficSign AI: Cyberpunk Neural Infrastructure 🚦🧠

A complete, production-ready, end-to-end AI-powered Traffic Sign Recognition System (TSRS) utilizing MobileNetV2 for high-speed, accurate classification. Features a stunning Cyberpunk-industrial dark mode UI, real-time webcam streaming, video analysis, and multi-lingual voice alerts (English & Hindi).

🚀 Key Features

• 🏎️ Ultra-Fast Inference: < 50ms per frame detection.
• 📸 Multi-Source Support: Images, Video Files, and Live Webcam.
• 🔊 Voice Synthesis: Automatic TTS alerts in English and Hindi.
• 📊 Neural Analytics: Real-time stats, detection trends, and distribution charts.
• 🌙 Night Mode: AI-enhanced visibility for low-light conditions.
• 🛡️ Admin Core: Secure panel for history management and system logs.
• 🐳 Docker Ready: Full containerization for instant deployment.

🛠️ Tech Stack

Component	Technology
Backend	Python / Flask / SQLAlchemy
Deep Learning	TensorFlow / Keras (MobileNetV2)
Computer Vision	OpenCV / NumPy
Frontend	Vanilla JS / CSS3 (Industrial Dark Mode)
Database	SQLite (Default) / PostgreSQL Compatible
Voice	gTTS / pyttsx3 (Offline Fallback)

📦 Quick Start (PowerShell)

1. Environment Setup

# Create virtual environment
python -m venv venv

# Activate virtual environment
.\venv\Scripts\Activate.ps1

# Install dependencies (with timeout for slow connections)
pip install --default-timeout=1000 -r requirements.txt

2. Dataset Initialization

# Auto-download GTSRB dataset and begin training
python train.py

3. Launch Neural Grid

# Start the web server
python app.py

Visit http://localhost:5000 in your browser.

🐳 Docker Deployment

Deploy instantly using Docker Compose:

docker-compose up --build

🧠 Model Specifications

• Architecture: MobileNetV2 with custom GlobalAveragePooling2D head.
• Classes: 43 (GTSRB Standard).
• Image Size: 64x64 pixels.
• Optimization: Adam optimizer, EarlyStopping, ReduceLROnPlateau.
• Accuracy Goal: 95%+ on GTSRB test set.

📁 Project Structure

TrafficSignAI/
├── dataset/            # GTSRB Data
├── models/             # Trained .h5 weights
├── static/             # Assets (CSS, JS, Audio)
├── templates/          # Jinja2 HTML Templates
├── utils/              # Preprocessing, DB, Voice logic
├── app.py              # Main Flask Entry
├── train.py            # Training Pipeline
└── detect.py           # Inference Engine

📡 API Documentation

Endpoint	Method	Description
/api/detect/image	POST	Upload image for analysis
/api/detect/video	POST	Start background video processing
/api/webcam/stream	GET	MJPEG real-time detection stream
/api/analytics/data	GET	Fetch system stats JSON
/api/admin/logs	GET	Paginated system logs

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Built with 💚 for Autonomous Infrastructure.