Expression Detection Using Flask

Project Plan

• Find a proper Face Expression detection model.
• Take live video from webcam with OpenCV.
• Use OpenCV haarcascades for face detection.
• Use ML model to predict expression.
• Integrate the model with the program.
• Build flask prototype.
• Make sure the flask app is running properly locally.
• Write client side javascript to take frames from the video to send to the server for detection.
• Handle the server response and draw on canvas proper bounding boxes.
• Deploy the flask app online for initial testing. (Heroku, done, working properly with automatic deployments on master)
• d3.js for visualizing a simple barchart.
• Design a Good looking UI. (Mostly completed)
• A big no for mobile devices.
• Find & Integrate Text Mood/Emotion Detection model.
• Write tests, move to TDD

Current todo:

• Block/black screen on devices.
• Try to fix the camera resolution in webrtc constraints.
• Search for new problems.

Setup

• Clone the git repo.
• Unzip the file.
• Navigate inside and open a CMD/Powershell Terminal.
• Create a virtual env using
  • Windows: python -m venv myvirtualenv
  • Linux: python3 -m venv myvirtualenv
• Activate virtual environment
  • Windows:
    • Powershell - myvirtualenv\Scripts\activate.ps1
    • CMD - myvirtualenv\Scripts\activate.bat
  • Linux: source myvirtualenv/bin/activate
• Install required packages
  • Windows: pip install -r requirements.txt
  • Linux: pip3 install -r requirements.txt
• Make sure to have latest Chrome/Firefox Installed.
• To run,
  • Windows: python server.py
  • Linux: python3 server.py

Deployment to heroku

• Create a new app.
• Connect your github repo to the app, select a branch and enable Automatic deployments.
• Install your OS Heroku CLI and login.
• Add a buildpack to the app to install required libraries for Open CV.

(env) kogam22@HOME-PC:~/code/FaceExpressionDetectionFlask$ heroku buildpacks:add --index 1 heroku-community/apt -a <your-app-name-here>
Buildpack added. Next release on <your-app-name-here> will use:
1. heroku-community/apt
2. heroku/python
Run git push heroku master to create a new release using these buildpacks.

• Update your branch to create new release.

Issues

Any problems ? Feel free to open up a new issue.