ChakmaLipī Dataset

This dataset contains 37,708 handwritten character images of the Chakma language, an endangered indigenous language of the Chittagong Hill Tracts in Bangladesh. The collection covers all 48 graphemes of the Chakma script, including 38 basic characters (vowels and consonants) and 10 numerals (0-9). Specifications: Format: .png images Dimensions: 40 x 40 pixels Channels: 3 (RGB) Total Classes: 48 Total Samples: 37,708 Directory Structure: The root folder contains 48 sub-directories, each corresponding to a specific class label.