I am using AWS Textract to fetch data from forms that I am uploading to S3. I am unable to find any method to perform this on documents in Japanese.
Reading through the AWS Docs shows that there is no official support.
Are there any workarounds?
Reading through the AWS Docs shows that there is no official support.
Correct. Currently, as per docs, Amazon Textract only supports English, Spanish, German, Italian, French, and Portuguese.
It doesn't support Japanese.
Are there any workarounds?
No, you need to use another OCR service that supports Japanese - either self-deployed, via the AWS Marketplace or another cloud provider.
A few self-deployed / AWS marketplace options could be Tesseract or NHocr.
Alternative options available on other cloud providers, in no specific order:
Azure AI Vision with OCR, optimised via the Document Intelligence read model, with full Japanese support for handwritten & print text
Google Cloud Vision OCR with full Japanese support for handwritten & print text
I would personally recommend just trying out Azure & GCP for your Japanese OCR needs & checking their suitability against your requirements e.g. cost, compliance, quality, ease of use etc.