Ancient Egyptian Hieroglyphs Segmentation and Classification with Convolutional Neural Networks

Abstract

Deep Learning is expanding in every domain of knowledge, allowing specialists to build tools to support their work in fields apparently unrelated to information technology. In this study, we exploit this opportunity by focusing on ancient Egyptian hieroglyphic texts and inscriptions. We investigate the ability of several convolutional neural networks (CNNs) to segment glyphs and classify images of ancient Egyptian hieroglyphs derived from various image datasets. Three well-known CNN architectures (ResNet-50, Inception-v3, and Xception) were considered for classification and trained on the supplied pictures using both the transfer learning and training from scratch paradigms. Furthermore, we constructed a specifically devoted CNN, termed Glyphnet, by changing the architecture of one of the prior networks and customizing its complexity to our classification goal. The suggested Glyphnet outperformed the others in terms of performance, ease of training, and computational savings, as judged by established measures. The ancient hieroglyphs segmentation was faced in parallel, using a deep neural network architecture known as Mask-RCNN. This network was trained to segment the glyphs, identifying the bounding box, which will be the input to a network for classification. Even though we focus here on single hieroglyph segmentation and classification tasks, the application of Deep Learning techniques in the Egyptological field opens up new and beneficial opportunities. In this light, the proposed work can be viewed as a jumping-off point for much more complex goals such as hieroglyphic sign coding, recognition, and transliteration; toposyntax of hieroglyphic signs combined to form words; linguistics analysis of hieroglyphic texts; recognition of corrupt, rewritten, and erased signs, and even identification of the scribe’s “hand” or sculptor’s school. This work shows how the ancient Egyptian hieroglyphs identification task can be supported by the Deep. Learning paradigm, laying the foundation for developing novel information tools for automatic documents recognition, classification and, most importantly, the language translation task.