2
[Feature] Add Baidu/Unlimited-OCR as an OCR option
Source: immich-app/immich#30624 · opened by @artamrj
I have searched the existing feature requests, both open and closed, to make sure this is not a duplicate request.
• Yes
The feature
I came across Baidu Unlimited-OCR today and it seems to be getting a lot of attention recently.
It is a new open-source (MIT) 3B OCR model with multilingual support and is designed for efficient long-context/document parsing.
If it fits the same use case as the OCR currently used in Immich, I think it could be interesting for the technical team to evaluate it as a possible way to improve OCR quality, especially for multilingual text.
Would be interested to hear what the Immich team thinks about it.
Platform
• Server
• Web
• Mobile
• Yes
The feature
I came across Baidu Unlimited-OCR today and it seems to be getting a lot of attention recently.
It is a new open-source (MIT) 3B OCR model with multilingual support and is designed for efficient long-context/document parsing.
If it fits the same use case as the OCR currently used in Immich, I think it could be interesting for the technical team to evaluate it as a possible way to improve OCR quality, especially for multilingual text.
Would be interested to hear what the Immich team thinks about it.
Platform
• Server
• Web
• Mobile
No pledges yet. Be the first to back this.
Comments
Similar requests
[Feature] Allow re-running OCR for a single asset
1 vote · 0 comments
[Feature] Translate OCR recognized text to target language using machine learning
2 votes · 0 comments
[Feature] Multi Model OCR
1 vote · 0 comments
[Feature] OCR hw acceleration for Rockchip boards
2 votes · 0 comments
[Feature]: MNN for Immich ML — OCR on Rockchip first; optional general backend?
2 votes · 0 comments
No comments yet.