In this chapter, we're going to solve one of the most challenging problems that arises from computer vision—real-time object detection. 

We'll begin by getting familiar with the object localization problem, and learn how to label data and modify the network layer of the neural network to fit this problem. Throughout this chapter, we'll learn how to detect an object with using the sliding window technique and the convolutional window technique. We'll conclude each section by exploring the downsides of the method and how to overcome them.  We'll take a closer look at the state-of-the-art algorithm, You Only Look Once (YOLO), that does a great job of detecting objects. We'll conclude this chapter by building a real-time application for object detection. 

In this chapter, we'll cover the following topics: 

• Resolving object localization
• Object detection using the sliding window algorithm
• Convolutional sliding window 
• Detecting objects using the YOLO algorithm
• Max suppression and anchor boxes 
• Building a real-time object detector in Java