This repository contains material and document related to the Wiedergutmachung project.
The Text Separation subfolder in this repository contains WGM-SYN, an artificial dataset for training and testing pixel-wise separation of machine-printed and handwritten text in historical archival documents. To evaluate the wgm_mixed.h5, run test_classifier.py with the test/ folder of the dataset WGM-SYN which can be downloaded from this repository.