YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Project Overview
This respository was created as part for an internship report, containing the results of the evaluation of MinerU document parsing model in content extraction from PDF documents. The evaluation assess the performance of MinerU on different document types and measures its ability in extracting document elements such as texts, tables, images, and overall document structure.
Objectives
The evaluation aims to:
- Assess the document parsing accuracy of MinerU
- Analyze the accuracy and precision of the results
- Identify strengths and limitations
- Compare parsing results across different document types
Installation
pip install --upgrade pip
pip install uv
uv pip install -U "mineru[all]"
How it works
GPU:
mineru -p -o
CPU:
mineru -p -o -b pipeline
Requirements
- PDF documents
- Python 3.10+
- MinerU v3.0.0 (At the time it was used)
Evaluation Workflow
- Prepare the PDF documents
- Run the MinerU to parse the PDFs
- Compare parsed outputs with the original pdf
- Review the model's performance and accuracy
Results
Please refer to the summary.md file for a more detailed summary.