Skip to content

README.md

Latest commit

 

History

History
executable file
·
255 lines (178 loc) · 4.79 KB

README.md

File metadata and controls

executable file
·
255 lines (178 loc) · 4.79 KB

🧠🔐 SentinelX AI

Autonomous Predictive Security Intelligence System

A full-stack AI-powered security platform that detects, analyzes, predicts, and responds to threats in real time with a centralized monitoring interface.

🚨 Problem Statement

Traditional surveillance systems are passive and reactive. They rely heavily on human monitoring, leading to:

  • Missed threats
  • Delayed responses
  • Lack of behavioral understanding
  • No predictive intelligence

SentinelX AI solves this by creating an autonomous, real-time AI security ecosystem.

⚡ Key Features

🧠 AI Threat Detection

  • Real-time object detection (YOLOv8)
  • Human and suspicious object detection
  • Bounding box visualization with confidence scores

🎯 Object Tracking

  • Multi-object tracking (DeepSORT)
  • Unique ID assignment per individual
  • Movement trajectory tracking

📊 Behavioral Intelligence

  • Feature extraction:

    • Speed
    • Direction
    • Dwell time

  • Anomaly detection (Isolation Forest / LSTM)

  • Detects:

    • Loitering
    • Sudden movement
    • Suspicious patterns

🧑‍💻 Face Recognition

  • Known vs unknown identification
  • Identity database integration
  • Blacklist alert system

🔮 Predictive Threat Engine

  • Time-series + probabilistic modeling
  • Risk scoring system
  • Future threat estimation

⚙️ Autonomous Decision Engine

  • Threat-based actions:

    • Monitor
    • Alert
    • Lockdown

🚨 Real-Time Alerts

  • WebSocket-based live alerts
  • Severity levels (Low / Medium / High)
  • Timestamp + metadata

🖥️ Monitoring Dashboard

A real-time AI Command Center Interface:

  • 📡 Live camera feeds with overlays
  • 📊 Activity timeline
  • 🚨 Alerts panel
  • 🔥 Threat heatmaps
  • 🧍 Identity tracking panel
  • 🧠 AI explainability panel
  • ⚙️ System health monitoring

🏗️ System Architecture

Frontend (React / Next.js)
        │
        ▼
WebSocket + REST API (FastAPI)
        │
        ▼
AI Pipeline
 ├── Detection (YOLOv8)
 ├── Tracking (DeepSORT)
 ├── Behavior Analysis
 ├── Face Recognition
        │
        ▼
Decision Engine → Alerts → Dashboard
        │
        ▼
Database (PostgreSQL + Redis)

⚙️ Tech Stack

Layer Technology
Backend FastAPI (Python)
AI Models PyTorch, OpenCV, YOLOv8
Frontend React / Next.js + TailwindCSS
Database PostgreSQL, Redis
Streaming WebRTC / RTSP
Deployment Docker, Docker Compose

📁 Project Structure

sentinelx-ai/
│
├── backend/
│   ├── ai/
│   ├── routes/
│   ├── services/
│   ├── models/
│   └── core/
│
├── frontend/
│   ├── components/
│   ├── pages/
│
├── docker-compose.yml
├── requirements.txt

🚀 Installation & Setup

1️⃣ Clone Repository

git clone https://github.com/yourusername/sentinelx-ai.git
cd sentinelx-ai

2️⃣ Setup Backend

cd backend
pip install -r requirements.txt
uvicorn main:app --reload

3️⃣ Setup Frontend

cd frontend
npm install
npm run dev

4️⃣ Run with Docker (Recommended)

docker-compose up --build

🔌 API Endpoints

Endpoint Description
/video/start Start video processing
/alerts Get all alerts
/system/status System health
/ws WebSocket for real-time updates

📡 WebSocket Events

  • new_alert
  • detection_update
  • system_metrics

🔐 Security & Privacy

  • End-to-end encryption
  • Role-based access control (RBAC)
  • Optional face anonymization
  • Secure data storage

⚡ Advanced Capabilities

  • Multi-camera support
  • Real-time processing pipeline
  • Scalable microservices architecture
  • Edge AI compatibility (Jetson Nano)
  • Federated learning (future scope)

🧪 Future Roadmap

  • 🔮 Predictive simulation engine (timeline forecasting)
  • 🌐 Digital twin environment
  • 🧠 Self-learning AI models
  • 📱 Mobile monitoring app

🤝 Contributing

Pull requests are welcome. For major changes, open an issue first to discuss your ideas.

📜 License

MIT License

👁️ Vision

SentinelX AI is designed to evolve into a fully autonomous security intelligence system, capable of predicting and preventing threats before they happen—transforming surveillance into intelligent defense.

⭐ Support

If you find this project useful:

  • ⭐ Star the repo
  • 🍴 Fork it
  • 🚀 Build on top of it