Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Building a Voice-Enabled Security Camera with a Mobile App and a Raspberry Pi

Building a Voice-Enabled Security Camera with a Mobile App and a Raspberry Pi

Linuxing in London, December 20th, 2017

In this session we’ll use deep learning powered services from AWS, to build a security camera with face and object recognition, configurable via a text or voice chatbot interface in a mobile app, using the AWS IoT platform to provide a secure and scalable connection. Updated with new Amazon Rekognition features and AWS DeepLens.

Danilo Poccia

December 20, 2017
Tweet

More Decks by Danilo Poccia

Other Decks in Programming

Transcript

  1. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
    Building a Voice-Enabled Security Camera
    with a Mobile App and a Raspberry Pi
    @danilop
    danilop
    [email protected]

    View Slide

  2. DIY Security Camera

    View Slide

  3. Raspberry Pi Zero W + Camera Module

    View Slide

  4. Control the Security Camera
    with a Mobile Phone
    Using Text or Voice

    View Slide

  5. How can I make my device “see”?

    View Slide

  6. Object and Scene Detection
    Image Moderation
    Facial Analysis
    Face Comparison
    Facial Recognition
    Celebrity Recognition
    A
    m
    azon
    R
    ekognition

    View Slide

  7. A
    m
    azon
    R
    ekognition

    View Slide

  8. A
    m
    azon
    R
    ekognition

    View Slide

  9. A
    m
    azon
    R
    ekognition

    View Slide

  10. A
    m
    azon
    R
    ekognition

    View Slide

  11. A
    m
    azon
    R
    ekognition

    View Slide

  12. A
    m
    azon
    R
    ekognition

    View Slide

  13. A
    m
    azon
    R
    ekognition

    View Slide

  14. A
    m
    azon
    R
    ekognition

    View Slide

  15. A
    m
    azon
    R
    ekognition
    Video

    View Slide

  16. How can I make my device “speak”?

    View Slide

  17. A
    m
    azon
    Polly
    A service that turns text
    into lifelike speech
    50 voices
    24 languages

    View Slide

  18. Content Creation
    Example: Convert an article to speech
    and download as MP3
    A
    m
    azon
    Polly

    View Slide

  19. Mobile and Desktop Apps
    You can store standard speech
    responses on the device, and also
    enable dynamic, real-time responses
    such as in-game character dialog,
    leaderboard rankings, and game
    invitations
    A
    m
    azon
    Polly

    View Slide

  20. Language Learning
    A
    m
    azon
    Polly

    View Slide

  21. How can I make my device understand “speech”?

    View Slide

  22. Conversational interfaces for your applications
    Powered by the same deep learning technologies as Alexa
    A
    m
    azon
    Lex
    Automatic Speech Recognition (ASR)
    for converting speech to text
    Natural Language Understanding (NLU)
    to recognize the intent of the text

    View Slide

  23. Informational Bots
    Build an Amazon Lex bot that allows
    patients to book appointments
    A
    m
    azon
    Lex

    View Slide

  24. Application Bots
    Get your banking information through
    an Amazon Lex chatbot
    A
    m
    azon
    Lex

    View Slide

  25. Internet of Things (IoT)
    Get your banking information through
    an Amazon Lex chatbot
    A
    m
    azon
    Lex

    View Slide

  26. How can I keep my IoT devices securely connected?

    View Slide

  27. AW
    S
    IoT

    View Slide

  28. And the final result is…

    View Slide

  29. AWS IoT
    S3 Bucket
    Amazon
    Rekognition
    Raspberry Pi
    Zero W +
    Camera Module
    Lambda
    Function
    Amazon
    Lex
    A
    rchitecture
    Smartphone
    Export
    Web Browser

    View Slide

  30. C
    ase
    Study
    Y-cam Protect
    A full-featured, home-security
    alarm system with wireless
    access and motion sensors

    View Slide

  31. C
    ase
    Study
    Y-cam Protect
    Using Amazon Polly to provide a human-sounding voice

    View Slide

  32. What about Developers?

    View Slide

  33. AW
    S
    D
    eepLens
    The world’s first
    deep learning enabled
    video camera
    for developers

    View Slide

  34. AW
    S
    D
    eepLens

    View Slide

  35. AW
    S
    D
    eepLens
    How AWS DeepLens Works

    View Slide

  36. Go Build!

    View Slide

  37. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
    Building a Voice-Enabled Security Camera
    with a Mobile App and a Raspberry Pi
    @danilop
    danilop
    [email protected]

    View Slide