Illustration: Artificial intelligence at work.
Russian specialists have developed an advanced artificial intelligence technology capable of translating ancient Egyptian hieroglyphs. This AI model was trained on the most comprehensive collection of ancient Egyptian texts, which includes their existing translations, as reported by the Artificial Intelligence Research Institute (AIRI).
Precision and Training
A collaborative team of researchers from the AIRI Institute, ISP RAS (Institute for System Programming of the Russian Academy of Sciences), and ITMO University successfully created a system for accurately recognizing and translating hieroglyphic inscriptions. The primary data source for training the model was the «Thesaurus Linguae Aegyptiae,» recognized as the most extensive compilation of ancient Egyptian texts alongside their translations. The model`s accuracy was rigorously assessed in two distinct stages: initially through automated metrics, and subsequently validated by the expert evaluation of professional Egyptologists from HSE University.
Wide-Ranging Applications
This innovative development offers a wide range of practical applications, notably within augmented reality mobile applications. Museum visitors will soon be able to point their smartphone cameras at ancient Egyptian artifacts and instantly receive on-screen translations of the inscriptions. Furthermore, this AI promises to significantly accelerate and enhance the work of researchers in deciphering and interpreting newly discovered materials. It is also poised to become an invaluable educational tool for Egyptology students who are just beginning their journey into studying the ancient language.
Tackling Complexity
Unlike standard alphabets, ancient Egyptian hieroglyphs present a significantly greater challenge for automated recognition. The AI system must navigate and interpret hundreds of unique symbols, many of which bear visual similarities, thereby complicating the decipherment process. The successful development of this system represents a major leap in overcoming these inherent complexities.
Dr. Ilya Makarov, a leading research fellow at the AIRI Institute and project leader, commented on the achievement: «This project is the result of a collaborative effort by an interdisciplinary team comprising machine learning specialists and Egyptologists. Our initial scientific paper detailing the development methodology has already been presented at the international SIGGRAPH 2025 conference. Currently, access to the system is available upon request for interested specialists.»
