PaperStructure

PaperStructure is a lightweight CLI tool designed to transform academic papers into clean, structured Markdown. By leveraging ONNX models, it delivers high-performance inference optimized for standard laptops. It is a reliable companion for formula-heavy research, though users may currently observe lower accuracy in table recognition.

Features

• Layout Detection -- YOLOX detects titles, sections, paragraphs, formulas, tables, figures
• Text Recognition -- PP-OCRv5 ONNX pipeline
• Formula Recognition -- Encoder-decoder LaTeX OCR
• Markdown Export -- clean, readable markdown output
• Parallel Processing -- multi-threaded PDF page processing

Demo

PDF	Markdown

Installation

pip install paper-structure

This registers the paper-structure CLI and installs the Python package.

CLI Usage

# Process a PDF (full pipeline: layout + OCR + formula)
paper-structure process paper.pdf -o output.md

# Shorthand:
paper-structure paper.pdf -o output.md

# OCR an image (text recognition, no layout detection)
paper-structure process photo.png -o output.txt

# Recognize a formula image as LaTeX
paper-structure process formula.png --formula

# PDF options
paper-structure process paper.pdf --max-pages 5 -v --save-images

# Generate annotated preview PDF with bounding boxes
paper-structure preview paper.pdf -o preview.pdf

# Manage models
paper-structure models status
paper-structure models download

Python API

PDF processing (full pipeline)

from paper_structure import PaperStructurePipeline

pipeline = PaperStructurePipeline()
result = pipeline.process_pdf("paper.pdf")
print(result["markdown"])
pipeline.save_markdown(result, "output.md")

Image OCR

from paper_structure import OCR

ocr = OCR()

# Text recognition (default)
print(ocr("table.png"))

# LaTeX formula recognition
print(ocr("formula.png", formula=True))

Model Management

from paper_structure.models import registry

registry.ensure_all()       # pre-download everything
print(registry.status())    # show cache status

Models

The tool automatically downloads models on its first call. All model weights are hosted at hpllduck/PaperStructure (~399 MB total) and cached locally via huggingface_hub.

Group	Files	Description
latex_ocr	encoder, decoder, image_resizer, tokenizer	RapidLaTeXOCR formula recognition
yolox	yolox_l0.05.onnx	YOLOX-L document layout detection
paddle_ocr	det, cls, rec, dictionary	PP-OCRv5 text detection/recognition

License

Apache License 2.0. Individual model weights retain their original licenses (MIT for LaTeX OCR, Apache-2.0 for YOLOX and PaddleOCR).

Acknowledgments

• PaddleOCR -- text recognition
• OnnxOCR -- ONNX OCR pipeline
• RapidLaTeXOCR -- formula recognition
• unstructured-inference -- layout detection