PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition_v2 #3497
Replies: 16 comments 12 replies
-
模型微调链接失效了吗 |
Beta Was this translation helpful? Give feedback.
-
通用OCR产线 |
Beta Was this translation helpful? Give feedback.
-
通用OCR产线的多语言调用服务示例什么时候支持C#版本,有加入C#版本的计划吗? |
Beta Was this translation helpful? Give feedback.
-
我注意到OCR中有很多不同的产线,哪里有页面介绍它们的不同吗? |
Beta Was this translation helpful? Give feedback.
-
如果识别的是pdf文件,最多识别10页吗,超过的部分怎么办呢 |
Beta Was this translation helpful? Give feedback.
-
有一个问题对于一行有多个单元格的情况然后每个单元格有多行文本的情况识别出来对于换行识别出错概率很大 |
Beta Was this translation helpful? Give feedback.
-
通用表格识别V2产线名字换了么:Exception: The pipeline (table_recognition_v2) does not exist! Please use a pipeline name or a config file path! |
Beta Was this translation helpful? Give feedback.
-
类似于这样的表格,识别出来的结果会是标题1、a、b、标题2、标题3、c、标题4、d,就是会有一个结果不是从上到下、从左到右的。这个问题该怎么处理啊 |
Beta Was this translation helpful? Give feedback.
-
我是个初学者,电脑账户是中文的情况下,很不友好 pipeline如何指定本地呢 |
Beta Was this translation helpful? Give feedback.
-
只要图片中有表格就报错:UnboundLocalError: cannot access local variable 'table_cells_pred' where it is not associated with a value怎么解决呢? |
Beta Was this translation helpful? Give feedback.
-
最新的paddle-gpu 和paddlex运行表格v2产线的结果比3.0rc0版本的差很多,我不知道这是为什么。还有可以将表格v2产线的文本识别和检测模型替换成最新的v5版本吗? |
Beta Was this translation helpful? Give feedback.
-
对于表格结构识别的结果,只有可视化的结果吗,如果我想要得到识别出来的每个单元格(包括空白的单元格)的坐标信息,应该如何处理 |
Beta Was this translation helpful? Give feedback.
-
为什么对于有些图片,可以产生表格结构识别结果(包括.xlsx文件,.json文件中的cell_box_list键,.html文件以及相应的table_cell_img.png文件),而对于别的图片则无法生成。我使用的方式为2.1命令行方式体验,使用的命令为paddlex --pipeline table_recognition_v2 |
Beta Was this translation helpful? Give feedback.
-
在进行表格识别v2产线时,从生成的table_cell_img图片来看,识别的都很正确,但是生成的html表格却是错乱的,这可能是什么原因?其中的哪个模型导致的? |
Beta Was this translation helpful? Give feedback.
-
--use_table_orientation_classify=False 参数无效 (img2table) PS C:\MyProgram\workspace\img2table> paddlex --pipeline table_recognition_v2 --use_doc_orientation_classify=False --use_doc_unwarping=False --use_table_orientation_classify=False --input table4.jpg --save_path ./output --device cpu |
Beta Was this translation helpful? Give feedback.
-
我是50显卡,驱动版本576.88 CUDA版本12.9,paddlex版本是3.2.0。 [notice] A new release of pip is available: 24.0 -> 25.2 我看官方说是支持了cuda12的呀,请问下这问题能解决吗,是没有适配50显卡吗 |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition_v2
https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition_v2.html
Beta Was this translation helpful? Give feedback.
All reactions