From self-driving cars to content moderation, detecting objects and their position in an image is a canonical task in computer vision. In this chapter, we will introduce techniques used for object detection. We will detail the architecture of two of the most prevalent models among the current state of the art—You Only Look Once (YOLO) and Regions with Convolutional Neural Networks (R-CNN).
The following topics will be covered in this chapter:
- The history of object detection techniques
- The main object detection approaches
- Implementing fast object detection using YOLO architecture
- Improving object detection using Faster R-CNN architecture
- Using Faster R-CNN with the TensorFlow Object Detection API
