How to handle complex merged cells in tables

[aneeqahsan01]

Hi, I have a pdf with complex merged cells and would like to know if there is some script (post processing) or any other means to handle this.
An example is this in pdf:

[aneeqahsan01]

@UserWangZz @jzhang533

[GreatV]

please refer to: https://github.com/PaddlePaddle/PaddleOCR/blob/main/ppstructure/docs/quickstart_en.md#pp-structure-quick-start

[aneeqahsan01]

That does not help a lot, The result is not a table just recognition of entities and texts. Here's the result:

I started exploring this: https://github.com/PaddlePaddle/PaddleOCR/blob/main/ppstructure/table/README.md
But the results are very bad.

[UserWangZz]

you need use your data finetune the table recognition model.

[saloni644c]

Hey, can you elaborate on finetune part? I am having the same issues with such tables