我不知道图片中的语言，希望ocr检测出文本和对应语种的概率，有这个模型吗？

PaddlePaddle / PaddleOCR

Awesome multilingual OCR toolkits based on PaddlePaddle (practical ultra lightweight OCR system, support 80+ languages recognition, provide data annotation and synthesis tools, support training and deployment among server, mobile, embedded and IoT devices)

Apache License 2.0

38.99k stars 7.32k forks source link

我不知道图片中的语言，希望ocr检测出文本和对应语种的概率，有这个模型吗？ #12021

Closed AnitaSherry closed 2 weeks ago

AnitaSherry commented 2 weeks ago

我有看到Multilingual_PP-OCRv3_det_infer模型，但是使用时也必须要指定语言才行

tink2123 commented 2 weeks ago

可以通过训练一个语种分类器，作为OCR的前置操作

AnitaSherry commented 2 weeks ago

能给个提示吗，在huggingface没搜到这样的模型

TachibanaYoshino commented 2 weeks ago

能给个提示吗，在huggingface没搜到这样的模型

Hi, I believe this work can help you, I have changed the DB code in paddleocr2.7 according to the tutorial given in the blog post, now it can classify text while detecting it. For example, the detected text can be classified into different languages, and the text, formulas, images, and tables in the picture can be detected and classified.

UserWangZz commented 2 weeks ago

感谢讨论❤️