From the course: Process Documents with Azure AI Document Intelligence

Unlock this course with a free trial

Join today to access over 25,500 courses taught by industry experts.

Solution: Creating a custom model

Solution: Creating a custom model

From the course: Process Documents with Azure AI Document Intelligence

Start my 1-month free trial Buy for my team

Solution: Creating a custom model

“

(exciting music) - [Instructor] So the goal of our challenge was to create a custom model to capture the required fields in the COVID consent forms and to display the results using the studio that rest API and Python. Here's how we accomplished the task. We first created a storage account and saved our training data in a block container called exercise files, in my case. We then labeled all the documents by assigning the proper fields. The key thing to note here is that one, some forms didn't have the middle name, so I had to draw a region for the empty fields. Two, the gender needed three selection marks. And three, the receipt vaccine checkbox also needed two selection marks. After we finished labeling, we created a neural model, which in my case is called LinkedInExercise02. The reason we chose a neural model for the exercise was because I tried it with a template model and the results were not as good. We then test the neural model we created. We uploaded the test file in the…

Contents