Katakana_Written_Character_Model

This handwritten recognition dataset was manually created for academic and research purposes. The dataset contains 85 distinct classes consisting of both handwritten characters and words. The samples