IIIT5K dataset converted to coco format along with python readable original label files. Original dataset is in matlab format, which might have been an issue for some potential users, hence this repository.
python converter ocr dataset text-recognition coco object-detection alphanumeric-strings number-detection character-detection alphanumeric-characters letter-detection
-
Updated
Mar 12, 2023