Skip to main content
The quality of data capture can be improved by training ABBYY FlexiCapture for Invoices Cloud. In a stream of documents of the same type, some unusual documents may have field locations that differ from those on a standard document. If the application cannot reliably detect fields on such documents, you can indicate the correct field locations, and the application “learns” to detect these fields. If a dataset of business units and vendors is connected to your application, ABBYY FlexiCapture for Invoices Cloud trains the capture model by default after the data has been verified. To train the capture model manually:
1

Run a processing task and verify the results

Run a document processing task and verify the results. For details, see Extraction of main invoice fields and Verification of extracted data.
2

Get the data layout

After the task status changes to Done, get the data layout using the Training.Layout method.
The response contains the data layout, including each detected field and its coordinates:
3

Train the data capture model

Provide the corrected field coordinates using the Training.Train method.
If the training completes successfully, the server returns a 204 No Content HTTP status.