This book will appeal to three broad groups of people. The first are intermediate iOS developers who are interested in learning and applying machine learning (ML); some exposure to ML concepts may be beneficial but are not essential as this book covers the intuition behind the concepts and models used throughout it.

The second group are those who have experience in ML but not in iOS development and are looking for a resource to help them to get the grips with Core ML; for this group, it is recommended to complement this book with a book that covers the fundamentals of iOS development. 

The last group are experienced iOS developers and ML practitioners who are curious to see how various models have been applied in the context of the iOS platform. 

Chapter 1, Introduction to Machine Learning, provides a brief introduction to ML, including some explanation of the core concepts, the types of problems, algorithms, and general workflow of creating and using a ML models. The chapter concludes by exploring some examples where ML is being applied.

Chapter 2, Introduction to Apple Core ML, introduces Core ML, discussing what it is, what it is not, and the general workflow for using it.

Chapter 3, Recognizing Objects in the World, walks through building a Core ML application from start to finish. By the end of the chapter, we would have been through the whole process of obtaining a model, importing it into the project, and making use of it.

Chapter 4, Emotion Detection with CNNs, explores the possibilities of computers understanding us better, specifically our mood. We start by building our intuition of how ML can learn to infer your mood, and then put this to practice by building an application that does just that. We also use this as an opportunity to introduce the Vision framework and see how it complements Core ML. 

Chapter 5, Locating Objects in the World, goes beyond recognizing a single object to being able to recognize and locate multiple objects within a single image through object detection. After building our understanding of how it works, we move on to applying it to a visual search application that filters not only by object but also by composition of objects. In this chapter, we'll also get an opportunity to extend Core ML by implementing customer layers. 

Chapter 6, Creating Art with Style Transfer, uncovers the secrets behind the popular photo effects application, Prisma. We start by discussing how a model can be taught to differentiate between the style and content of an image, and then go on to build a version of  Prisma that applies a style from one image to another. We wrap up this chapter by looking at ways to optimize the model. 

Chapter 7, Assisted Drawing with CNNs, walks through building an application that can recognize a users sketch using the same concepts that have been introduced in previous chapters. Once what the user is trying to sketch has been recognized, we look at how we can find similar substitutes using the feature vectors from a CNN. 

Chapter 8, Assisted Drawing with RNNs, builds on the previous chapter and explores replacing the the convolution neural network (CNN) with a recurrent neural network (RNN) for sketch classification, thus introducing RNNs and showing how they can be applied to images. Along with a discussion on learning sequences, we will also delve into the details of how to download and compile Core ML models remotely. 

Chapter 9, Object Segmentation Using CNNs, walks through building an ActionShot photography application. And in doing so, we introduce another model and accompanying concepts, and get some hands-on experience of preparing and processing data.

Chapter 10, An Introduction to Create ML, is the last chapter. We introduce Create ML, a framework for creating and training Core ML models within Xcode using Swift. By the end of this chapter, you will know how to quickly create, train, and deploy a custom models. 

To be able to follow through the examples in this book, you will need the following software:

For the examples that are dependent on Core ML 2, you will need the following software:

It's recommended that you use https://notebooks.azure.com (or some other Jupyter notebook service provider) to follow the examples using the Core ML Tools Python package, but those wanting to run locally or train their model will need the following software:

public protocol VideoCaptureDelegate: class {
    func onFrameCaptured(
      videoCapture: VideoCapture, 
      pixelBuffer:CVPixelBuffer?, 
      timestamp:CMTime)
}

@IBOutlet var previewView:CapturePreviewView!
@IBOutlet var classifiedLabel:UILabel!

let videoCapture : VideoCapture = VideoCapture()

Feedback from our readers is always welcome.

General feedback: Email [email protected] and mention the book title in the subject of your message. If you have questions about any aspect of this book, please email us at [email protected].

Errata: Although we have taken every care to ensure the accuracy of our content, mistakes do happen. If you have found a mistake in this book, we would be grateful if you would report this to us. Please visit www.packtpub.com/submit-errata, selecting your book, clicking on the Errata Submission Form link, and entering the details.

Piracy: If you come across any illegal copies of our works in any form on the Internet, we would be grateful if you would provide us with the location address or website name. Please contact us at [email protected] with a link to the material.

If you are interested in becoming an author: If there is a topic that you have expertise in and you are interested in either writing or contributing to a book, please visit authors.packtpub.com.