Object detection goes one step further than the image classification discussed in the previous chapter. Image classification just returns a label of class for an image, while object detection returns a list of objects identified in an image, along with a bounding box for each identified object. Modern object detection algorithms use deep learning to build models that can be used to detect and localize all kinds of objects in a single image. In the past few years, faster and more accurate object detection algorithms came one after another, and in June 2017, Google released the TensorFlow Object Detection API that incorporates several leading object detection algorithms.

In this chapter, we’ll first give a quick overview of object detection: the process of creating an effective deep learning model for object detection and then using the...