[Feature] Translate OCR recognized text to target language using machine learning

[Sesshoumaru-sama]

I have searched the existing feature requests, both open and closed, to make sure this is not a duplicate request.

• Yes

The feature

Hi

The adding of OCR to Immich is a great thing, but for me to be actually usefull I would need a translation feature as well.

Background:
I have many anime images which feature japanese text. The recognition works fine. However, as I am not too "fuent" in japanese, I cannot use it for searching. And I also want to see the text translated on the image (maybe via toggle original/translation).

Basically I would need something like Google Lense, so the feature request would be:

The workflow would be OCR recognition, then translation from recognized language to configured target language, then using original text and translation for search index.

On the image itself the option to show original OCR text and via toggle switch to translated text, maybe a config which of both is default

Find all images with OCR test on them, maybe by adding tag "OCR" to all that images

I didn not see any other discussion about this, maybe I overlooked it. But I think this would be really helpful. Especially if it would be done via machine leanring locally (not some cloud).

Best Regards

Platform

• Server
• Web
• Mobile