Digital Assyriology—Advances in Visual Cuneiform Analysis

Cuneiform tablets appertain to the oldest textual artifacts used for more than three millennia and are comparable in amount relevance texts written Latin or ancient Greek. These typically found Middle East were by imprinting wedge-shaped impressions into wet clay. There is an increasing demand Digital Humanities domain handwriting recognition, i.e., machine reading of handwritten script, focusing on historic documents. Current practice text analysis cuneiform script relies heavily transliteration translation, which incomplete influenced knowledge experience expert that created them. The development computational tools presents many opportunities. An efficient accurate sign spotting enables cross-referencing statistical analyzes infeasible perform manually. Furthermore, a wedge constellation tool, provides experts with significantly broader base references create less time consuming transliterations translations. Yet, has since resisted efforts processing basis its basic constituents, 3D their free-form arrangements signs. In this work, we review literature recognition script. We introduce different heterogeneous sources namely, manual ink-on-paper drawings, digital vector graphics photographs, scans tablets. describe methods, beginning first classification using hybrid encoding comparison, latest methods making use Generative Adversarial Neural Networks recognize characters automatically. Finally, give overview applications these enable quantitative mining small, e.g., patterns constellations, large, networks economic activity.