In this chapter, we covered how we can use image processing using a popular deep-learning-based API service. We also discussed how we can apply the same with a custom trained model, by extending a previously created base model. While we did not explicitly mention it, the extension of the base model was a part of the process termed transfer learning (TL), where models trained on a certain dataset are imported into and used in a completely different scenario, with little or minimal fine-tuning.
Furthermore, the chapter covered why and when TensorFlow Lite is a good candidate for building a model, and how Flutter can be used for applying the same on the device model, which runs offline and is very fast. This chapter sets a milestone, with the introduction of Python and TensorFlow into the project, both of which will be used extensively in the upcoming chapters.
In the...