This repository contains source code used in an ongoing research on automatic extraction of the reading order of digitized handwritten text documents. For details see our paper
Along with the source code, splits for training/validation and test are defined for three different dataset, namely:
- Download data:
./get_data.sh - Train and test the models:
./run.sh