ML5 Feature Extractor

This sample application is created using ml5js library.

ml5.js is machine learning for the web in your web browser. It is built on the well-known-one-and-only TensorFlow.js.

Installation

git@github.com:yong-asial/ml5-feature-extractor.git
cd ml5-feature-extractor
npm install

Preview

cd ml5-feature-extractor
npm run dev

How the program is working

• The program (DOM) is fully loaded.
• It will load the core model (MobileNet).
• Now the program is ready to either load the existing custom model or train a new classification.

Load custom model

• Click on load button. It will then load the custom model located in the model directory. It is the custom model detecting whether you are wearing mask or not.
• Click on Start Detecting! button - You can try facing to your camera by wearing mask or not and observe the Result label.

Train new classification

• Input the class name in the input box Class 1 and Class 2.
• Show something to the camera (for example, not wearning mask) and click Class 1 button.
• Do the samething but this time show something different to camera (for example, wearing mask) and click Class 2 button
• Try to differentiate your pose or angle (or anything). Try to add image more than 20 times or so.
• Click Train and observe the loss. Once the training is completed, you can click on Start Detecting! button