Ubisoft using machine learning to create 'ultimate hieroglyphs translator'

Ubisoft wants to use the power of machine learning to simplify the decipherment of ancient hieroglyphics.

The studio realized how "challenging it still is to unlock the secrets of the past" during the development of  Assassin's Creed: Origins, and kick started 'The Hieroglyphics Initiative' to dissect one of the oldest written languages with the help of modern day tools. 

The research project is powered by Google TensorFlow, and will use image recognition techniques to analyze pictures of hand-carved hieroglyphs in order to identify and highlight those symbols we already understand. 

In doing so, Ubisoft hopes to create the "ultimate hieroglyphs translator" by building a library of source material based on the thousand or so hieroglyphs that have already been identified and translated by scholars in the past. 

"Translating hieroglyphics is a manual and time-consuming process. It has been over 200 years since the Rosetta Stone first helped to understand them," wrote the French studio. 

"The ancient Egyptians used hieroglyphics for over 3000 years. We’re focusing on hieroglyphics from the Middle Egyptian language, which was used from 2000 BC until as late as the 4th Century AD.

"Because of their age, the quality of inscriptions carved on stone three to four thousand years ago varies greatly – with many having missing or damaged portions. It’s our belief that machine learning can transform the process of collating, cataloguing and understanding the written language of the Pharaohs."

Ubisoft plans to share its first set of results by the end of the year, and is calling for historians and researchers to get involved with the project. You can find out how to take part over on The Hieroglyphics Initiative website.