[Question]: about LayoutRecognizer4YOLOv10 labels

[Danee-wawawa]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (Language Policy).
• Non-english title submitions will be closed directly ( 非英文标题的提交将会被直接关闭 ) (Language Policy).
• Please do not modify this template :) and fill in all the required fields.

Describe your problem

Hi, thank you for your work.
I have a question, why do the Figure caption and Table caption have to be repeated here?

https://github.com/infiniflow/ragflow/blob/main/deepdoc/vision/layout_recognizer.py#L178

[dosubot]

The repetition of "Figure caption" and "Table caption" in the labels list of the LayoutRecognizer4YOLOv10 class appears to be intentional in the code, but there is no explicit documentation or comments explaining the rationale behind this design choice. The labels list is as follows:

labels = [
    "title",
    "Text",
    "Reference",
    "Figure",
    "Figure caption",
    "Table",
    "Table caption",
    "Table caption",
    "Equation",
    "Figure caption",
]

This means both "Figure caption" and "Table caption" appear more than once in the list. Based on the available code and documentation, there is no clear explanation for why this repetition is necessary or what specific purpose it serves for the model or recognition tasks. It could be related to model requirements, label alignment with a specific dataset, or a workaround for a particular use case, but this is not documented in the codebase.

If you need a definitive answer, it may be helpful to reach out to the maintainers or authors of this part of the code for clarification, as the intent is not clear from the current code or documentation [source].

To continue talking to Dosu, mention @dosu-bot.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}

[Danee-wawawa]

someone can answer?