PaddleOCR Finnish Language Support Inquiry

[iey704]

I am writing to inquire about Finnish language OCR support and would appreciate your guidance on the following matters:

Key Questions

1. Finnish Recognition Model Update Schedule

• Could you please share the approximate timeline for Finnish recognition (rec) model updates? We would greatly appreciate any information regarding planned improvements or releases.

2. Compatibility with Current Environment

• Our project is currently running on the following versions:

paddleocr==2.10.0
paddlepaddle==2.6.2

• We would like to know if Finnish language support improvements can be applied to our current setup without requiring updates to these library versions, or if version upgrades will be necessary.

Additional Context

• We are actively working on a project that requires robust Finnish text recognition capabilities, and understanding the development roadmap would help us plan our implementation strategy accordingly.

We look forward to your response and appreciate all the excellent work the PaddleOCR team continues to do.
Thank you.

[liuhongen1234567]

Hello, PPOCRv5 has added support for the Finnish language in version 3.1.0. You can perform Finnish language recognition by setting lang="fr".

Here is an example code:

from paddleocr import PaddleOCR

ocr = PaddleOCR(
    lang="fr", # Specify French recognition model with the lang parameter
    use_doc_orientation_classify=False, # Disable document orientation classification model
    use_doc_unwarping=False, # Disable text image unwarping model
    use_textline_orientation=False, # Disable text line orientation classification model
)
result = ocr.predict("https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_french01.png")
for res in result:
    res.print()
    res.save_to_img("output")
    res.save_to_json("output")

To utilize the French recognition capabilities of PP-OCRv5, you need to upgrade the paddleocr version to at least 3.1.0. If you want to use paddleocr==2.10.0, you can use PP-OCRv3's Latin language recognition. However, there may be some differences in recognition accuracy compared to v5.

For more details, please refer to the documentation. https://paddlepaddle.github.io/PaddleOCR/main/en/version3.x/algorithm/PP-OCRv5/PP-OCRv5_multi_languages.html#2-quick-start

[iey704]

Hello, first of all, thank you for your response.
However, I have some questions that have arisen, so I am writing this reply. Looking at the parameters applicable to the official documentation language, I can see that lang=fr corresponds to French. I need to confirm once more whether the Finnish language also falls under this category.
I am curious whether applying 'fr' results in a Latin language model that includes Finnish being applied, which is why you provided the response as shown above.
I will await your response.

[liuhongen1234567]

Hello, due to the large number of languages in the world, creating a model for each language is very challenging. Therefore, we have chosen to train models based on language families rather than individual languages. According to our research, French belongs to the Latin language family (also known as the Romance language group). As a result, we use a Latin language family model to recognize a series of languages such as English, French, German, etc.