In this chapter, we started from the very simple intuition behind the task of object detection and then proceeded to very advanced concepts, such as Instance Segmentation, which is a contemporary research area. Object detection is at the heart of a lot of innovation in the field of Retail, Media, Social Media, Mobility, and Security; there is a lot of potential for using these technologies to create very impactful and profitable features for both enterprise and social consumption. 

From the Algorithms perspective, this chapter started with the legendary Viola-Jones algorithm and its underlying mechanisms, such as Haar Features and Cascading Classifiers. Using that intuition, we started exploring the world of CNN for object detection with algorithms, such as R-CNN, Fast R-CNN, up to the very state-of-the-art Faster R-CNN.

In this chapter, we also laid the foundations and introduced a very recent and impactful field of research called instance segmentation. We also covered some state-of-the-art Deep CNNs based on methods, such as Mask R-CNN, for easy and performant implementation of instance segmentation.