Victorian400: Colorizing Victorian Illustrations
Kim Ho-Yeol · International Journal of Humanities and Arts Computing · 2021
This article introduces the Victorian400 dataset, created to colorize black-and-white illustrations from nineteenth-century British novels with deep learning models. The Victorian400 dataset is a set of colour illustrations painted using nineteenth-century palettes which have been shared as open sources. This article examines the process of creating and curating the Victorian400 dataset and the validity of the Victorian400 dataset by looking into the results of the test set with the trained set based on conditional generative adversarial networks.