Face Mask Detection using ML.NET Model Builder C#

Blog: https://code-ai.mk
Tutorial:

Today we are building a Face Mask Detection using C#. This demo is created using Accord.NET Haar Object Classifier to detect faces and then ML.NET Model Builder to create and train mask detection module. Both applications are connected via CMD Service that enable us to use ML.NET model in a .NET Framework application.
This project allows you to choose an image. Set face detection parameters and by clicking the "Detect" button it will display face bounding boxes. Each bounding box can have 2 colors. Green if the face on the image wears a face mask, red otherwise.
First the app checks if the user has selected a valid image. Then it pre-processes the image and passes it to the Face Detection module. This component will return an array of rectangles (bounding boxes). These bounding boxes are then used to crop out the face image. Finally each detected face is passed to the ML.NET application for classification. The image can belong to one of two categories. Either a person wears a mask or not. This result is represented in the main form of the application by changing the color to the rectangle accordingly.