DAS 2018 Tutorial on OCR and Layout Analysis

This repository contains a number of worksheets and presentations that show how to use the collection of PyTorch-based OCRopus components.

The presentations are numbered, so it's recommended that you read the in order.

If you want to run these examples (or simply install all the OCR components), you can do that with:

pip install -r requirements.txt

You also need to install PyTorch and CUDA.

Name		Name	Last commit message	Last commit date
Latest commit History 10 Commits
figs		figs
pipeline		pipeline
testdata		testdata
.gitattributes		.gitattributes
10-overview.ipynb		10-overview.ipynb
20-pytorch.ipynb		20-pytorch.ipynb
25-dlinputs.ipynb		25-dlinputs.ipynb
30-full-pipeline.ipynb		30-full-pipeline.ipynb
40-augmentation.ipynb		40-augmentation.ipynb
50-intro-ocrobin.ipynb		50-intro-ocrobin.ipynb
60-intro-ocrorot.ipynb		60-intro-ocrorot.ipynb
70-lstm-layout.odp		70-lstm-layout.odp
70-lstm-layout.pdf		70-lstm-layout.pdf
75-intro-ocroseg.ipynb		75-intro-ocroseg.ipynb
80-lstm-text.odp		80-lstm-text.odp
80-lstm-text.pdf		80-lstm-text.pdf
85-intro-ocroline.ipynb		85-intro-ocroline.ipynb
99-TODO.ipynb		99-TODO.ipynb
LICENSE		LICENSE
README.md		README.md
helpers.py		helpers.py
helpers.pyc		helpers.pyc
ocropus3-recognize		ocropus3-recognize
requirements.txt		requirements.txt

Provide feedback