On recognition of Cyrillic Text

Kostiantyn Liepieshov, Oles Dobosevych

Keywords: Computer Vision, Optical Character Recognition, Transfer Learning, Datasets, Handwritten Text Recognition, Document Intelligence, Cyrillic Text

Abstract: We introduce the largest (among publicly available) dataset for Cyrillic Handwritten Text Recognition and the first dataset for Cyrillic Text in the Wild Recognition, as well as suggest a method for recognizing Cyrillic Handwritten Text and Text in the Wild. Based on this approach, we develop a system that can reduce the document processing time for one of the largest mathematical competitions in Ukraine by 12 days and the amount of used paper by 0.5 ton.

TL;DR: We introduce several datasets for Cyrillic OCR and a method for its recognition