/EmoDet

Emotion Detection with TensorFlow Light for RaspberryPi 4 using Python and Django

Primary LanguagePython

EmoDet

Emotion Detection with TensorFlow Light for RaspberryPi 4 using Python and Django

Usage

  • Depending on the available mobile device, in order to configure it as a Django Development Server that runs this software you will have to install the following packages and libraries:

    • django
    • numpy
    • opencv
    • tensorflow lite
    • pathlib
    • mss
    • pyautogui

    Most of them can be installed through PIP if you are using a RaspberryPI with ArchLinux or ManjaroLinux (aarch64 version). The Tensorflow Light wheel has to be built (to this date) if you want to use python3.8.

How to star a server with Django:

  • cd into EmoDet folder where manage.py is located
  • Disable DHCP on your connection with the router and set a static IP, for example: 192.168.1.170
  • Simply run python manage.py runserver 192.168.1.170:8000
  • If you want to use a different port just remember to add it to EmoDet/settings.py.

How to use the server:

  • Open a browser from any device on local network and type 192.168.1.170:8000
  • Choose the desired acquisition type
  • Enjoy