Machine Learning for iOS Developers

Chapter 5Object Detection Using Pre-trained Models

WHAT'S IN THIS CHAPTER

• Introduction to object detection
• Introduction to artificial neural networks
• Introduction to convolutional neural networks
• Using a pre-trained model to implement object detection in an iOS app

In this chapter, you will learn about the basics of artificial neural networks (ANNs) and convolutional neural networks (CNNs) and use a pre-trained convolutional neural network model with Core ML to create an app that allows users to take a picture to detect the dominant object in the image.

NOTE  You can download the code files for this chapter from wrox.com or from GitHub using the following URL:

https://github.com/asmtechnology/iosmlbook-chapter5.git

What Is Object Detection?

Before we jump into the code to implement object detection in an app, it will be useful to understand what object detection is and how it works. Object detection is an application of computer vision techniques to detect one or more objects in an image. Computer vision is a discipline within artificial intelligence that deals with the development of algorithms that can be applied to the domain of digital images and video.

Early applications of object detection algorithms were primarily used for quality control in manufacturing facilities. Typically, a still or video camera would be used to take an image of a part (such as a circuit board), and this image would be processed by a computer to detect any anomalies in the image. If ...