Chitti app logo

Chitti is better in the app

Coins, streaks and builds — free on Android & iOS

Get the app
Chitti logoChitti

Mask ON & OFF

4 Minutes · Module 4 - Image Processing

Teachable Machine mask detection training for students teaches how to build an AI model for mask on and mask off. A Tamil technology lesson.

  • Creating mask on and mask off classes in Teachable Machine
  • Recording webcam image samples and training the model
  • Testing the model with and without a face mask and saving it
What this lesson teachesRead the lesson as text

A mask detection model is a machine learning system trained to recognize whether a person is wearing a face mask or not. In automated entry systems, this type of computer vision model decides access by allowing people wearing masks to enter through a door while keeping the door closed for those without a mask.

Setting up the image project

Training begins by opening the Teachable Machine website and selecting a new standard image project. Machine learning systems learn by sorting visual examples into distinct categories known as classes. In this project, two separate classes are created: one labeled mask on for faces covered by a mask, and another labeled mask off for bare faces.

Collecting image samples

  • Image samples are captured directly through a webcam showing faces with and without masks.
  • When a webcam cannot be used, pre-saved image files can be downloaded and uploaded directly into the classes.
  • Gathering numerous pictures of different friends provides the system with varied examples of each state.
  • Loading a larger number of image samples requires more time for the computer to process during training.

Training and testing the model

Clicking the train model button starts the machine learning process on all the collected visual data. The software studies the differences between the images in the mask on class and the mask off class. Once training is complete, the model is tested in real time using the webcam. Facing the camera without a mask displays a mask off result, and putting a face mask on immediately changes the output to mask on.

Saving the finished model

Once the live test verifies that the predictions are working accurately, the project is ready to be exported. Selecting the download project as file option saves the finished machine learning model locally on the computer. This stored project file can then be integrated into future computer programs, such as custom websites or mobile applications.

Part of the course

A.I. with Teachable Machine

View course →

More in Module 4 - Image Processing

  1. 1Fruit7 Minutes
  2. 2Mask ON & OFFNow playing