The Handwritten Hebrew Dataset (HHD_v0)
Hebrew, handwritten, letters
HDD_v0 consists of images of isolated Hebrew characters together with training and test sets subdivision.
The images were collected from hand-filled forms.
For more details, please refer to .
When using this dataset in research work, please cite .
The dataset is divided into TRAIN and TEST set (folders), each one containing 27 subfolders.
Each subfolder contains the images of a letter from the alphabet (one subfolder for each letter of the alphabet).