Deep Learning is a great model for handling unstructured data, especially on images. The progress of this field is really fast, and one of the progress is something called Transfer Learning.

Transfer Learning is a method to train the neural network that has already trained on a different dataset, so we don’t have to train it from scratch because it could take several days or weeks to train them.

If we use the transfer learning to our dataset, it only takes several hours to train because we only train the final layer. Therefore, we can use it to train on the other dataset with already pre-trained model and its given architecture.

To make the model is useful to use, we have to deploy them, in example by building a web app that makes it more user friendly. Thankfully, we can do that using PyTorch to build a deep learning model and Flask to build a web application.

In this article, I will show you on how to build a web application for image classification on an Apple leaf to classify whether is it healthy or not and if it doesn’t, which disease the leaf has. To build that, we can use transfer learning using PyTorch, and also how to build a simple web application using Flask. Here is the preview of the web application,

#programming #machine-learning #deep-learning #data-science #flask