Masked Self-supervised Pre-training for Text Recognition Transformers on Large-Scale Datasets.

Martin Kiss, Michal Hradis

Browse the full ICDAR paper archive.