Detectron2 for Document Layout Analysis

📚 Process PDFs, Word documents and more with spaCy

Trained Detectron2 object detection models for document layout analysis based on PubLayNet dataset

自动给文本文件的中文英文之间加入合理的空格

Trained Detectron2 object detection models for document layout analysis based on PubLayNet dataset

Proof of concept of training a simple Region Classifier using PdfPig and ML.NET (LightGBM). The objective is to classify each text block in a pdf document page as either title, text, list, table and image.

A collection of machine learning apis in fastapi, can be deployed with uvicorn, hypercorn or docker

Enable the line numbering and implement the LayoutVisitor to count rows in a document.