Frigate

Frigate

Frigate monitors your security cameras with locally processed AI. It provides real-time object detection, person detection, and intelligent video analysis using machine learning models, all processed locally for privacy and performance.

Similar self-hosted alternatives:
Repository activity:
Stars
22,925
Forks
2,153
Watchers
198
Open Issues
150
Last commit
2 days ago
Details:
Estimated Popularity
96
Pricing Model
Free
Hosting Type
Self-Hosted
License
MIT
Deployment Difficulty
Medium
Language
TypeScript
Frigate screenshot

Frigate is a cutting-edge AI-powered video surveillance system that brings intelligent object detection to security cameras. By processing video streams locally with machine learning models, it provides real-time detection of people, vehicles, and other objects while maintaining privacy and reducing false alarms.

Key Features

  • Advanced AI Detection:

    • Real-time object detection using TensorFlow Lite
    • Person detection with high accuracy
    • Vehicle and animal detection
    • Custom object detection models
    • Configurable detection zones
    • False positive filtering
  • Intelligent Video Analysis:

    • Motion detection with AI validation
    • Zone-based detection areas
    • Time-based detection rules
    • Event correlation and filtering
    • Smart alert generation
    • Confidence threshold tuning
  • Camera Management:

    • RTSP stream support
    • Multiple camera integration
    • Hardware-accelerated decoding
    • Stream quality optimization
    • Camera health monitoring
    • Flexible camera configuration
  • Event Management:

    • Automatic event recording
    • Intelligent clip creation
    • Snapshot generation
    • Event timeline visualization
    • Search and filtering capabilities
    • Retention policy management
  • Smart Notifications:

    • Real-time alert system
    • Configurable notification rules
    • MQTT message publishing
    • Webhook integration
    • Mobile push notifications
    • Email alerts
  • Home Automation Integration:

    • Native Home Assistant integration
    • MQTT broker support
    • RESTful API access
    • Custom automation triggers
    • State reporting
    • Sensor data publishing
  • Performance Optimization:

    • Hardware acceleration support
    • GPU processing (NVIDIA, Intel)
    • Coral TPU integration
    • CPU optimization
    • Memory management
    • Efficient storage usage

Technical Specifications

  • Language: Python
  • AI Framework: TensorFlow Lite
  • Deployment: Docker containers
  • Protocols: RTSP, MQTT, HTTP
  • Hardware: CPU, GPU, Coral TPU
  • License: MIT
  • Platforms: Linux, Docker

Use Cases

  • Smart Home Security: Intelligent home surveillance with AI detection
  • Business Monitoring: Commercial security with reduced false alarms
  • Perimeter Security: Outdoor monitoring with vehicle and person detection
  • Pet Monitoring: Animal detection and behavior monitoring
  • Package Detection: Delivery and package monitoring
  • Privacy-Focused Surveillance: Local processing without cloud dependency

Unique Advantages

  • Local AI Processing: All detection happens locally for privacy and speed
  • Intelligent Filtering: AI reduces false alarms from motion detection
  • Hardware Acceleration: Support for various AI acceleration hardware
  • Home Assistant Native: Deep integration with popular home automation platform
  • Open Source: Transparent, customizable, and community-driven
  • Real-time Performance: Low-latency detection and alerting

Based on the Frigate GitHub repository, this tool revolutionizes home and business security by bringing enterprise-grade AI detection capabilities to self-hosted surveillance systems, making it ideal for users who want intelligent monitoring without compromising privacy or relying on cloud services.

Help improve this content

Found an error or want to add more information about Frigate? You can edit this page directly on GitHub.

Project Categories

Click on a category to explore similar projects