Handwritten Babylonian Numerals (HBN) Dataset

HBN Dataset Description The Handwritten Babylonian Numerals (HBN) Dataset is an openly accessible collection of 14,000 original handwritten images of Babylonian numeral symbols collected from 100 participants at the Higher Colleges of Technology, United Arab Emirates. The dataset comprises 14 numeral classes representing the values 1–9, 10, 20, 30, 40, and 50. Each participant reproduced each numeral class ten times using a structured handwriting collection form, producing 140 images per participant and 1,000 original images per class. The completed forms were scanned at 300 dpi. Individual numeral samples were extracted using predefined coordinates, printed collection frames were removed, and the images were converted to grayscale and standardized to 128 × 128 pixels. The dataset is organized by numeral class. Structured filenames preserve anonymized participant identifiers, numeral values, and repetition indices, enabling reproducible experiments and writer-independent evaluation. The HBN Dataset supports research in handwritten numeral recognition, computer vision, deep learning, image classification, structural feature analysis, historical document processing, computational archaeology, and digital humanities. The collection consists of modern handwritten reproductions of Babylonian numerals rather than images of archaeological clay tablets. The released dataset contains 14,000 original handwritten samples; augmented or synthetically generated images are not included. The dataset is released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Related GitHub repository: https://github.com/drloay40/HBN-Dataset Copy description 2. License and copyright Creative Commons Attribution 4.0 International Copyright © 2026 Loay Alzubaidi.