Skip to content
A tutorial on the PyTorch-based ocropus components.
Jupyter Notebook Python
Branch: master
Clone or download
Pull request Compare This branch is even with tmbdev:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
figs
pipeline
testdata
10-overview.ipynb
20-pytorch.ipynb
25-dlinputs.ipynb
30-full-pipeline.ipynb
40-augmentation.ipynb
50-intro-ocrobin.ipynb
60-intro-ocrorot.ipynb
70-lstm-layout.odp
70-lstm-layout.pdf
75-intro-ocroseg.ipynb
80-lstm-text.odp
80-lstm-text.pdf
85-intro-ocroline.ipynb
99-TODO.ipynb
LICENSE
README.md
helpers.py
helpers.pyc
ocropus3-recognize
requirements.txt

README.md

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.

You can’t perform that action at this time.