Charlas - Monica Limachi: Proyecto Centinela: Seguridad en el hogar usando Raspberry Pi y Tensorflow

Learn how to build an affordable home security system using Raspberry Pi and TensorFlow that can identify household members and detect unknown subjects through AI-powered monitoring.

Key takeaways
  • Project Centinela focuses on creating a smart home security system using Raspberry Pi and TensorFlow

  • The system uses cameras and object detection to identify known and unknown subjects in the home environment

  • Key components include:

    • Raspberry Pi hardware
    • Small surveillance cameras
    • Pre-trained models like Mobile Net V2 and Efficient Net
    • Label Box for image labeling/annotation
  • The workflow involves:

    • Collecting and storing photos/images
    • Labeling/tagging subjects (household members, pets)
    • Training the model to recognize authorized people
    • Detecting and alerting about unknown subjects
  • The project aims to create an affordable, personal security solution rather than using expensive enterprise systems

  • Uses open-source tools and pre-trained models to make implementation more accessible

  • System is designed to store and process images locally for privacy/security

  • The solution automatically identifies and classifies subjects entering the monitored space

  • Focus is on practical implementation for home use rather than large-scale deployment

  • Demonstrates how AI/ML technologies can be applied to personal security needs using accessible hardware