{"id":1088,"date":"2018-07-24T11:56:37","date_gmt":"2018-07-24T09:56:37","guid":{"rendered":"https:\/\/zaven.co\/blog\/?p=1088"},"modified":"2025-04-08T19:55:08","modified_gmt":"2025-04-08T17:55:08","slug":"core-ml-machine-learning-for-ios","status":"publish","type":"post","link":"https:\/\/zaven.co\/blog\/core-ml-machine-learning-for-ios\/","title":{"rendered":"Core ML: Machine Learning for iOS"},"content":{"rendered":"

You have probably heard phrases like \u201cMachine Learning\u201d or \u201cArtificial Intelligence\u201d without knowing exactly what they mean. In this article, I will shed some light on \u201cWhat Machine Learning is\u201d by showing you a Core ML iOS sample<\/strong>.<\/p>\n

It’s really hard to explain \u201cWhat Machine Learning is<\/em>\u201d in one sentence because nowadays it is a sprawling field. In my opinion, the easiest way to reply to that question is: learning from experience.<\/p>\n

What is Machine Learning?<\/h2>\n

In computer science, Machine Learning means that a computer is able to learn from experience in order to understand data<\/em> (for example; images, text, number or almost anything that we could express as binary data).<\/p>\n

Equally important as data is an algorithm, which tells the computer how to learn from that data.<\/strong> After applying that algorithm to the data on a powerful computer, we could get an artifact that is created by the training process as a result. That artifact can be exported to lightweights executable – trained model.<\/p>\n

What is a machine learning model?<\/h3>\n

Apple Inc. introduced Core ML framework in 2017<\/strong> which provides functionality to integrate trained machine learning models into an iOS or a macOS app.<\/p>\n

\"machine

Apple Core ML model. Source: medium.com<\/p><\/div>\n

As mobile developers, we don’t need to be an expert in machine learning at all. We just want to play with machine learning features, without wasting time on creating our own trained models.<\/p>\n

Fortunately, you could find hundreds of trained models on the internet. For example from Caffe Model Zoo<\/a>, TensorFlow Models<\/a> or MXNet Model Zoo<\/a>. That easily converted to Core ML model format (models with a .mlmodel<\/code> file extension)  using conversion tools like mxnet-to-coreml<\/code> or tfcoreml<\/code>.<\/p>\n

You can also get a model from Apple Developer<\/a> which works out of the box.<\/p>\n

Let\u2019s get through the process of integrating a Core ML model into your app<\/strong> step by step.<\/p>\n

Core ML – Machine Learning iOS<\/h2>\n

I\u2019ve created a sample project which shows how to use a machine learning model on iOS 11 using Core ML<\/em>.<\/p>\n

For this purpose, I\u2019ve used Vision – a framework for image analysis, which is built on top of Core ML, and the Pixabay API<\/a>, that allows us to download photos.<\/p>\n

\"core

Core ML model. Source: developer.apple.com<\/p><\/div>\n

The idea is simple:<\/strong><\/p>\n

    \n
  1. Download pictures from Pixabay (thanks to the Kingfisher<\/a> library for downloading and caching images from the web)<\/li>\n
  2. Predict where a photo was taken based on the view from that photo<\/li>\n<\/ol>\n

    The first part of the work is pretty easy: make the API call using URLSession<\/code>, parse it and display the results in UICollectionView<\/code>.<\/p>\n

    Once we have got our image downloaded, we can make a prediction. First, we need to create our VNCoreMLModel<\/code>,  whose base class is a container for a Core ML model with Vision requests (for that app I have used RN1015k500<\/code>).<\/p>\n

    After that, we need to initialize our request object, which is a type of VNCoreMLRequest<\/code>. To do that, we need our previously initialized object from the VNCoreMLModel<\/code>. The request object is initialized with completion closure which returns VNRequest<\/code> or Error<\/code>.<\/p>\n

    The last object that we need to create for our prediction is VNImageRequestHandler<\/code>, which accepts data types like:<\/p>\n