Azure Document Intelligence Layout Model - Word Doc issues

Question

Hello, the layout model still not supports the selection mark extraction and skips images in the word document.

Are any enhancements planned for release ?

Thanks

Answer 1

Hi there Devang Grewal

Thanks for using QandA platform

Yeah thats right the Layout model in Azure Document Intelligence does not support selection mark extraction (like checkboxes) or inline image extraction from Word documents.

There hasn’t been any officially announced a specific timeline for these features in .docx support. You can keep an eye on the Azure AI updates page for roadmap changes.

As a workaround Convert Word docs to PDF before processing this enables more accurate layout and image extraction.

If this helps kindly accept the answer thanks much.

Answer 2

Hi Devang Grewal

The Azure Document Intelligence Layout Model is designed to analyze documents and extract structured data, which includes understanding the layout of headers, tables, and other elements. While converting a DOCX file to PDF might be necessary for certain workflows, it can lead to a loss of some structural advantages that the DOCX format provides. The layout model excels in capturing the relationships and continuations of elements like headers and tables when working directly with DOCX files, as it can leverage the inherent structure of the document.

If maintaining the structural integrity and layout of your document is crucial, it would be beneficial to use the DOCX format for analysis with the layout model instead of converting it to PDF.

Kindly refer below link: input-requirements

Thank You.