Can Azure Form Recognizer v3.1.1 read foreign language and currency characters?

Question

Hello,

I am currently testing out Azure Form Recognizer v3.1.1 to see how it handles non-english characters (i.e. Chinese) and currency characters (i.e. GBP, euros). I am using this quickstart program (https://learn.microsoft.com/en-us/azure/applied-ai-services/form-recognizer/quickstarts/get-started-sdk-rest-api?pivots=programming-language-csharp) and when I get the results back from recognizerClient, it is returning a "?" character for every chinese character found.

It is the same for currency symbols. While I am able to pick up GBP, I cannot pick up euros or rupees as a character.

Is this a version issue? The latest live version is 3.1.1, and this page here (https://learn.microsoft.com/en-us/azure/applied-ai-services/form-recognizer/language-support) states that not setting that specific language will allow for documents with multiple languages (I have been testing a pdf with both chinese and english characters). Please let me know if v3.1.1 is able to pick up foreign languages and currency symbols.

Thanks,

Answer 1

@Vincent Villacorta

Hello,

Yes, Form Recognizer support language as below list but only for "Layout" and "Custom Model" feature and only after V3.0. I just did a test of Chinese Simplified and traditional and it works well.

https://learn.microsoft.com/en-us/azure/applied-ai-services/form-recognizer/language-support

Hope this helps! Please kindly accept the answer if you feel helpful!

Regards,
Yutong