Transfer learning image classifier

In this tutorial, you will learn how to build a custom image classifier that you will train on the fly in the browser using TensorFlow.js.

You will use transfer learning to create a highly accurate model with minimal training data. You will be using a pre-trained model for image classification called MobileNet. You will train a model on top of this one to customize the image classes it recognizes.

This tutorial is presented as a codelab. Follow this link to open the codelab