In this chapter, we saw how to develop an end-to-end project that will detect objects from video frames when video clips play continuously. We saw how to utilize the pre-trained Tiny YOLO model, which is a smaller variant of the original YOLO v2 model.
Furthermore, we covered some typical challenges in object detection from both still images and videos, and how to solve them using bounding box and non-max suppression techniques. We learned how to process a video clip using the JavaCV library on top of DL4J. Finally, we saw some frequently asked questions that should be useful in implementing and extending this project.
In the next chapter, we will see how to develop anomaly detection, which is useful in fraud analytics in finance companies such as banks, and insurance and credit unions. It is an important task to grow the business. We will use unsupervised learning algorithms...
