[Feature] Addition of Layout Detection Algorithm

[liuwake]

What is the feature?

• I would like to request the implementation of a layout detection algorithm in the repository. This feature would greatly enhance the functionality and usability of the project for users who require such capabilities. Could we discuss the feasibility and potential timeline for this addition?

• Datasets: GROTOAP2, PubLayNet, DocBank, and CLDA.

• Algorithm:DocXchain, Surya360LayoutAnalysis, and LayoutLM.

Any other context?

No response