Summary
As you can see, these three algorithms make it easy to train CV models. Even with default hyperparameters, we get good results pretty quickly. Still, we start feeling the need to scale our training jobs. Don't worry: once the relevant features have been covered in future chapters, we'll revisit some of our CV examples and we'll scale them radically!
In this chapter, you learned about the image classification, object detection, and semantic segmentation algorithms. You also learned how to prepare datasets in image, RecordIO, and SageMaker Ground Truth formats. Labeling and preparing data is a critical step that takes a lot of work, and we covered it in great detail. Finally, you learned how to use the SageMaker SDK to train and deploy models with the three algorithms, as well as how to interpret results.
In the next chapter, you will learn how to use built-in algorithms for natural language processing.
