The corpus accounts for 22M OCRed characters along with the corresponding Gold Standard (GS).
Currently browsing:
Dataset Content Types
Scope
Languages
Licenses