Microsoft updates Azure Form Recognizer: Invoices go multi-language
Power Apps Express Design doesn't get to have all the AI fun
Days after the debut of doodle-recognizing Express Design on the Power Apps platform, Microsoft has updated its Azure sibling: Form Recognizer.
Azure Form Recognizer, as its name suggests, pulls text and structure from documents using AI and OCR. The theory goes that users can automate data processing with the tech, which accepts PDFs, scanned images and handwritten forms (although, as with all handwriting recognition systems, scrawl barely readable by humans can equally stump the robots.)
More usefully, Azure Form Recognizer can map field relationships as key-value pairs and spit out some structured JSON without "excessive manual intervention" as Microsoft delicately puts it.
The latest preview of the tech adds the ability to extract paragraphs – handy for unstructured documents – and roles for those paragraphs (for example: titles or footnotes.) The June 2022 API version will also spot tabular fields, useful for turning document content into tables. It will also handle tables that span multiple pages.
"If you have a dataset labeled with tables," explained Microsoft, "train a model with the current API to start seeing multi-page tables in the response."
Other tweaks include the ability to extract text from Word, Excel and Powerpoint files, as well as text from embedded images. HTML documents can also be scanned. US driver license scanning has been improved to extract fields including DateOfIssue, Height and Weight, and Japanese has been added to the business card model.
- Scribble to app: Microsoft's Power Apps VP talks us through 'Express design'
- House! Microsoft wins at buzzword bingo with AI, ML and Blockchain in cloudy Azure wrapping
- Google previews Document AI for parsing forms: Just a catch-up with AWS and Azure?
- Microsoft Azure to spin up AMD MI200 GPU clusters for 'large scale' AI training
However, it is the additional languages now supported by the invoice model that are likely the most attractive. Spanish and English are joined by German, French, Italian, Portuguese, and Dutch. "This opens up the procurement scenarios to invoices in many different languages," said Microsoft.
A caveat: we would caution readers to check for regulatory compliance with that data. This is all, after all, being processed in Azure, which might give some lawmakers pause for thought.
Still, for developers tasked with document digitization and riding the Azure train, the updates (in preview) will welcome. There is, however, unlikely to be enough to prise engineers from rival products by Google and AWS, such as Cloud Vision and Textract respectively. ®
- Active Directory
- Internet Explorer
- Microsoft 365
- Microsoft Build
- Microsoft Edge
- Microsoft Office
- Microsoft Surface
- Microsoft Teams
- Office 365
- Patch Tuesday
- SQL Server
- Visual Studio
- Visual Studio Code
- Windows 10
- Windows 11
- Windows 7
- Windows 8
- Windows Server
- Windows Server 2003
- Windows Server 2008
- Windows Server 2012
- Windows Server 2013
- Windows Server 2016
- Windows XP
- Xbox 360