Using Document AI to extract information from documents
As you may recall, we discussed Document AI at length in Chapter 3. We talked about how it goes beyond understanding the content of textual inputs and also incorporates structure. In this section, we’re going to take a look at how Document AI does this, and we’ll see it in action for some real-world use cases. We’ll start by covering some important Document AI concepts.
Document AI concepts
Document AI uses the concept of “processors” to process documents. There are three main types of processors that you can use:
- General processors
- Specialized processors
- Custom processors
General processors and specialized processors use models that are pre-trained by Google, so you can use them without needing to train your own models, and without any AI/ML expertise. However, to provide additional flexibility, Google Cloud gives you the ability to further train specialized processors...
