Chapter 6: Training Models with Visual Data

Deep learning has been successfully applied to many different types of data, including tabular data, text data, and recommender system data. You saw fastai's approach to these types of data in Chapter 3, Training Models with Tabular Data, Chapter 4, Training Models with Text Data, and Chapter 5, Training Recommender Systems. These types of data are all part of the story of deep learning, but visual data or image data is the type of data that is traditionally associated most closely with deep learning.

Visual data is also the type of data that is most thoroughly supported by the fastai framework. The fastai high-level API is mostly developed for visual data, and 70% of the curated fastai datasets are visual datasets. In this chapter, we will explore some of the features that fastai provides for exploring visual datasets and building high-performance deep learning models with image datasets.

In this chapter, you will learn how to...