Глобальный теггер

Abstract“Document Information Extraction (DIE) is a crucial task for extracting key information fromvisually-rich documents. The typical pipeline approach for this task involves Optical Charac-ter Recognition (OCR), serializer, Semantic Entity Recognition …...