CATMuS Modern and Contemporary (McCATMuS) – Ground Truth Dataset

MC-CATMuS is a dataset of ground truth for automatic text recognition aligned on the framework proposed by Consistent Approach for Transcribing Manuscripts (CATMuS). It focuses on the documents dating from the modern and contemporary (“MC”) periods (from 1600 to present days), written in languages using the Latin script, and mixing multiple types of writing: handwritten, printed or typewritten. It results from the aggregation of already existing ground truth datasets and allows usto investigate the feasibility of harmonizing transcription practices after the fact.